Datasets:
microtensor-io
/
agent-sessions

Dataset card Data Studio Files
xet
 Community
item_id
stringlengths
21
26
family
stringclasses
383 values
twin_of
stringlengths
0
21
system
stringlengths
1.51k
1.51k
user
stringlengths
18.6k
288k
input_tokens
int64
6.38k
123k
budget_tokens
int64
1k
3.07k
cut
int64
19
151
steps
int64
32
166
resolved
bool
1 class
image
stringlengths
25
33
source
stringclasses
1 value
facts
stringlengths
585
1.33k
probes
stringlengths
1.29k
1.44k
ctx_compress-s1-01608-twin
principalledger
ctx_compress-s1-01608
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] revoke_scope_all ignores the cap, so late attempts wait for minutes. It lives in src/principalledger/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the principalledger...
22,683
1,134
45
94
false
synth:8368186625043601183@1
synth
[{"aliases": ["src/principalledger/report.py", "report.py"], "kind": "file", "value": "src/principalledger/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["data/nonce_2.csv", "nonce_2.csv"], "kind": "file", "value": "data/nonce_2.csv"},...
[{"answers": ["src/principalledger/report.py", "report.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agen...
ctx_compress-s1-01609
corridordesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_tile raises KeyError for a missing key instead of returning the default. It lives in src/corridordesk/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the corridor...
25,265
1,263
54
61
false
synth:8135324038334651080@1
synth
[{"aliases": ["src/corridordesk/audit.py", "audit.py"], "kind": "file", "value": "src/corridordesk/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/corridordesk/report.py", "report.py"], "kind": "file", "value": "src/corridordesk/report...
[{"answers": ["src/corridordesk/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's m...
ctx_compress-s1-01609-twin
corridordesk
ctx_compress-s1-01609
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_tile raises KeyError for a missing key instead of returning the default. It lives in src/corridordesk/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the corridor...
14,566
1,000
34
61
false
synth:8135324038334651080@1
synth
[{"aliases": ["src/corridordesk/audit.py", "audit.py"], "kind": "file", "value": "src/corridordesk/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/corridordesk/report.py", "report.py"], "kind": "file", "value": "src/corridordesk/report...
[{"answers": ["src/corridordesk/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's m...
ctx_compress-s1-01610
tariffforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reconcile_receipt_page skips the first page; page 1 returns what should be page 2. It lives in src/tariffforge/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tariffforge...
36,038
1,801
78
85
false
synth:5644018386512517234@1
synth
[{"aliases": ["src/tariffforge/cli.py", "cli.py"], "kind": "file", "value": "src/tariffforge/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/tariffforge/core.py", "core.py"], "kind": "file", "value": "src/tariffforge/core.py"}, {"aliases": ["d...
[{"answers": ["src/tariffforge/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most r...
ctx_compress-s1-01610-twin
tariffforge
ctx_compress-s1-01610
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reconcile_receipt_page skips the first page; page 1 returns what should be page 2. It lives in src/tariffforge/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tariffforge...
20,833
1,041
46
85
false
synth:5644018386512517234@1
synth
[{"aliases": ["src/tariffforge/cli.py", "cli.py"], "kind": "file", "value": "src/tariffforge/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/tariffforge/core.py", "core.py"], "kind": "file", "value": "src/tariffforge/core.py"}, {"aliases": ["d...
[{"answers": ["src/tariffforge/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most r...
ctx_compress-s1-01611
warehouseforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_lot_all raises PalletShapeError for bytes input in strict mode instead of decoding it. It lives in src/warehouseforge/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by ma...
57,824
2,891
102
121
false
synth:6736363738318400461@1.42893
synth
[{"aliases": ["src/warehouseforge/report.py", "report.py"], "kind": "file", "value": "src/warehouseforge/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/warehouseforge/limits.py", "limits.py"], "kind": "file", "value": "src/warehou...
[{"answers": ["src/warehouseforge/report.py", "report.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent...
ctx_compress-s1-01611-twin
warehouseforge
ctx_compress-s1-01611
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_lot_all raises PalletShapeError for bytes input in strict mode instead of decoding it. It lives in src/warehouseforge/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by ma...
32,417
1,620
69
121
false
synth:6736363738318400461@1.42893
synth
[{"aliases": ["src/warehouseforge/report.py", "report.py"], "kind": "file", "value": "src/warehouseforge/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/warehouseforge/limits.py", "limits.py"], "kind": "file", "value": "src/warehou...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["io.py", "src/warehouseforge/io.py"], ...
ctx_compress-s1-01612
lexemedesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] tokenise_slug leaves a trailing dash when the title ends in punctuation. It lives in src/lexemedesk/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the lexemedesk layout....
55,761
2,788
110
125
false
synth:1641795666676811163@1.03205
synth
[{"aliases": ["src/lexemedesk/audit.py", "audit.py"], "kind": "file", "value": "src/lexemedesk/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/lexemedesk/compat.py", "compat.py"], "kind": "file", "value": "src/lexemedesk/compat.py"}, {...
[{"answers": ["src/lexemedesk/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mos...
ctx_compress-s1-01612-twin
lexemedesk
ctx_compress-s1-01612
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] tokenise_slug leaves a trailing dash when the title ends in punctuation. It lives in src/lexemedesk/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the lexemedesk layout....
26,972
1,348
58
125
false
synth:1641795666676811163@1.03205
synth
[{"aliases": ["src/lexemedesk/audit.py", "audit.py"], "kind": "file", "value": "src/lexemedesk/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/lexemedesk/compat.py", "compat.py"], "kind": "file", "value": "src/lexemedesk/compat.py"}, {...
[{"answers": ["src/lexemedesk/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mos...
ctx_compress-s1-01613
legforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] simplify_fix double counts carried amounts when the period closes on a weekend. It lives in src/legforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the legforge layo...
41,408
2,070
84
95
false
synth:6652512912321702822@1
synth
[{"aliases": ["src/legforge/rules.py", "rules.py"], "kind": "file", "value": "src/legforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases": ["da...
[{"answers": ["src/legforge/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01613-twin
legforge
ctx_compress-s1-01613
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] simplify_fix double counts carried amounts when the period closes on a weekend. It lives in src/legforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the legforge layo...
20,543
1,027
48
95
false
synth:6652512912321702822@1
synth
[{"aliases": ["src/legforge/rules.py", "rules.py"], "kind": "file", "value": "src/legforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases": ["da...
[{"answers": ["src/legforge/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01614
palletkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] restock_stock raises ValueError when a value itself contains an equals sign. It lives in src/palletkit/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the palletkit layou...
46,925
2,346
86
107
false
synth:2517038684914765645@1
synth
[{"aliases": ["src/palletkit/audit.py", "audit.py"], "kind": "file", "value": "src/palletkit/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["data/shipment_4.csv", "shipment_4.csv"], "kind": "file", "value": "data/shipment_4.csv"}, {"aliase...
[{"answers": ["src/palletkit/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most...
ctx_compress-s1-01614-twin
palletkit
ctx_compress-s1-01614
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] restock_stock raises ValueError when a value itself contains an equals sign. It lives in src/palletkit/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the palletkit layou...
26,495
1,324
58
107
false
synth:2517038684914765645@1
synth
[{"aliases": ["src/palletkit/audit.py", "audit.py"], "kind": "file", "value": "src/palletkit/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["data/shipment_4.csv", "shipment_4.csv"], "kind": "file", "value": "data/shipment_4.csv"}, {"aliase...
[{"answers": ["src/palletkit/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most...
ctx_compress-s1-01615
zoneledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] interpolate_fix_page raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/zoneledger/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
48,102
2,405
72
87
false
synth:8564702161240417244@1
synth
[{"aliases": ["src/zoneledger/compat.py", "compat.py"], "kind": "file", "value": "src/zoneledger/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/zoneledger/format.py", "format.py"], "kind": "file", "value": "src/zoneledger/format.p...
[{"answers": ["src/zoneledger/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01615-twin
zoneledger
ctx_compress-s1-01615
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] interpolate_fix_page raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/zoneledger/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
24,074
1,203
42
87
false
synth:8564702161240417244@1
synth
[{"aliases": ["src/zoneledger/compat.py", "compat.py"], "kind": "file", "value": "src/zoneledger/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/zoneledger/format.py", "format.py"], "kind": "file", "value": "src/zoneledger/format.p...
[{"answers": ["src/zoneledger/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01616
bookingforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] overlap_roster returns one item too few on every page. It lives in src/bookingforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the bookingforge layout. [tool_use kin...
37,127
1,856
84
103
false
synth:8976789884005043512@1.0441
synth
[{"aliases": ["src/bookingforge/rules.py", "rules.py"], "kind": "file", "value": "src/bookingforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/bookingforge/schema.py", "schema.py"], "kind": "file", "value": "src/bookingforge/schema...
[{"answers": ["src/bookingforge/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01616-twin
bookingforge
ctx_compress-s1-01616
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] overlap_roster returns one item too few on every page. It lives in src/bookingforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the bookingforge layout. [tool_use kin...
24,273
1,213
51
103
false
synth:8976789884005043512@1.0441
synth
[{"aliases": ["src/bookingforge/rules.py", "rules.py"], "kind": "file", "value": "src/bookingforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/bookingforge/schema.py", "schema.py"], "kind": "file", "value": "src/bookingforge/schema...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["rules.py", "src/bookingforge/rules.py...
ctx_compress-s1-01617
clauseworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_heading double counts carried amounts when the period closes on a weekend. It lives in src/clauseworks/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the clause...
90,814
3,072
130
139
false
synth:2032120996709542829@1.70528
synth
[{"aliases": ["src/clauseworks/compat.py", "compat.py"], "kind": "file", "value": "src/clauseworks/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/clauseworks/cli.py", "cli.py"], "kind": "file", "value": "src/clauseworks/cli.py"}, ...
[{"answers": ["src/clauseworks/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's ...
ctx_compress-s1-01617-twin
clauseworks
ctx_compress-s1-01617
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_heading double counts carried amounts when the period closes on a weekend. It lives in src/clauseworks/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the clause...
55,788
2,789
80
139
false
synth:2032120996709542829@1.70528
synth
[{"aliases": ["src/clauseworks/compat.py", "compat.py"], "kind": "file", "value": "src/clauseworks/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/clauseworks/cli.py", "cli.py"], "kind": "file", "value": "src/clauseworks/cli.py"}, ...
[{"answers": ["src/clauseworks/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's ...
ctx_compress-s1-01618
skuflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_bin misses keys that differ only in case. It lives in src/skuflow/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the skuflow layout. [tool_use kind=search name=Bash] fin...
59,951
2,997
84
93
false
synth:5749124393907363027@1
synth
[{"aliases": ["src/skuflow/cli.py", "cli.py"], "kind": "file", "value": "src/skuflow/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/sku_4.csv", "sku_4.csv"], "kind": "file", "value": "data/sku_4.csv"}, {"aliases": ["data/bin_5.csv", "bin_5.c...
[{"answers": ["src/skuflow/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most recen...
ctx_compress-s1-01618-twin
skuflow
ctx_compress-s1-01618
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_bin misses keys that differ only in case. It lives in src/skuflow/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the skuflow layout. [tool_use kind=search name=Bash] fin...
26,281
1,314
44
93
false
synth:5749124393907363027@1
synth
[{"aliases": ["src/skuflow/cli.py", "cli.py"], "kind": "file", "value": "src/skuflow/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/sku_4.csv", "sku_4.csv"], "kind": "file", "value": "data/sku_4.csv"}, {"aliases": ["data/bin_5.csv", "bin_5.c...
[{"answers": ["src/skuflow/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most recen...
ctx_compress-s1-01619
revisionkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_section returns the upper bound for every value instead of clamping into the range. It lives in src/revisionkit/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
100,184
3,072
116
139
false
synth:7451991457470490049@1.78665
synth
[{"aliases": ["src/revisionkit/limits.py", "limits.py"], "kind": "file", "value": "src/revisionkit/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/revision_9.csv", "revision_9.csv"], "kind": "file", "value": "data/revision_9.csv"}...
[{"answers": ["src/revisionkit/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's ...
ctx_compress-s1-01619-twin
revisionkit
ctx_compress-s1-01619
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_section returns the upper bound for every value instead of clamping into the range. It lives in src/revisionkit/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
57,884
2,894
72
139
false
synth:7451991457470490049@1.78665
synth
[{"aliases": ["src/revisionkit/limits.py", "limits.py"], "kind": "file", "value": "src/revisionkit/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/revision_9.csv", "revision_9.csv"], "kind": "file", "value": "data/revision_9.csv"}...
[{"answers": ["src/revisionkit/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's ...
ctx_compress-s1-01620
slotflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] assign_slot skips the first page; page 1 returns what should be page 2. It lives in src/slotflow/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the slotflow layout. [tool_use ...
63,333
3,072
84
103
false
synth:1152744257796366313@1.33957
synth
[{"aliases": ["src/slotflow/io.py", "io.py"], "kind": "file", "value": "src/slotflow/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/roster_1.csv", "roster_1.csv"], "kind": "file", "value": "data/roster_1.csv"}, {"aliases": ["data/calendar_4.csv"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_compat.py", "tests/test_compat.p...
ctx_compress-s1-01620-twin
slotflow
ctx_compress-s1-01620
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] assign_slot skips the first page; page 1 returns what should be page 2. It lives in src/slotflow/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the slotflow layout. [tool_use ...
30,564
1,528
49
103
false
synth:1152744257796366313@1.33957
synth
[{"aliases": ["src/slotflow/io.py", "io.py"], "kind": "file", "value": "src/slotflow/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/roster_1.csv", "roster_1.csv"], "kind": "file", "value": "data/roster_1.csv"}, {"aliases": ["data/calendar_4.csv"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_report.py", "tests/test_report.p...
ctx_compress-s1-01621
shipmentkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_reorder leaves a trailing dash when the title ends in punctuation. It lives in src/shipmentkit/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shipmentkit layout....
52,523
2,626
92
99
false
synth:8780878879617173059@1
synth
[{"aliases": ["src/shipmentkit/audit.py", "audit.py"], "kind": "file", "value": "src/shipmentkit/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/shipmentkit/rules.py", "rules.py"], "kind": "file", "value": "src/shipmentkit/rules.py"}, ...
[{"answers": ["src/shipmentkit/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mo...
ctx_compress-s1-01621-twin
shipmentkit
ctx_compress-s1-01621
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_reorder leaves a trailing dash when the title ends in punctuation. It lives in src/shipmentkit/audit.py; `python -m pytest tests/test_audit.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shipmentkit layout....
24,728
1,236
50
99
false
synth:8780878879617173059@1
synth
[{"aliases": ["src/shipmentkit/audit.py", "audit.py"], "kind": "file", "value": "src/shipmentkit/audit.py"}, {"aliases": ["tests/test_audit.py", "test_audit.py"], "kind": "file", "value": "tests/test_audit.py"}, {"aliases": ["src/shipmentkit/rules.py", "rules.py"], "kind": "file", "value": "src/shipmentkit/rules.py"}, ...
[{"answers": ["src/shipmentkit/audit.py", "audit.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mo...
ctx_compress-s1-01622
warehousedesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] pick_bin grows linearly instead of doubling. It lives in src/warehousedesk/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the warehousedesk layout. [tool_use kind=search n...
77,604
3,072
110
125
false
synth:5713043592502535097@2.10555
synth
[{"aliases": ["src/warehousedesk/core.py", "core.py"], "kind": "file", "value": "src/warehousedesk/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/warehousedesk/io.py", "io.py"], "kind": "file", "value": "src/warehousedesk/io.py"}, {"alias...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["data/reorder_7.csv", "reorder_7.csv"]...
ctx_compress-s1-01622-twin
warehousedesk
ctx_compress-s1-01622
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] pick_bin grows linearly instead of doubling. It lives in src/warehousedesk/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the warehousedesk layout. [tool_use kind=search n...
43,708
2,185
71
125
false
synth:5713043592502535097@2.10555
synth
[{"aliases": ["src/warehousedesk/core.py", "core.py"], "kind": "file", "value": "src/warehousedesk/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/warehousedesk/io.py", "io.py"], "kind": "file", "value": "src/warehousedesk/io.py"}, {"alias...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_rules.py", "tests/test_rules.py"...
ctx_compress-s1-01623
sampleledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] clip_probe misses keys that differ only in case. It lives in src/sampleledger/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sampleledger layout. [tool_use kind=search n...
24,534
1,226
69
74
false
synth:8846326222135599535@1
synth
[{"aliases": ["src/sampleledger/cli.py", "cli.py"], "kind": "file", "value": "src/sampleledger/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/probe_2.csv", "probe_2.csv"], "kind": "file", "value": "data/probe_2.csv"}, {"aliases": ["src/sampl...
[{"answers": ["src/sampleledger/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01623-twin
sampleledger
ctx_compress-s1-01623
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] clip_probe misses keys that differ only in case. It lives in src/sampleledger/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sampleledger layout. [tool_use kind=search n...
16,283
1,000
41
74
false
synth:8846326222135599535@1
synth
[{"aliases": ["src/sampleledger/cli.py", "cli.py"], "kind": "file", "value": "src/sampleledger/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/probe_2.csv", "probe_2.csv"], "kind": "file", "value": "data/probe_2.csv"}, {"aliases": ["src/sampl...
[{"answers": ["src/sampleledger/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01624
shiftdesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] split_window double counts carried amounts when the period closes on a weekend. It lives in src/shiftdesk/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shiftdesk ...
55,052
2,752
87
108
false
synth:4476294910221231004@1.00388
synth
[{"aliases": ["src/shiftdesk/limits.py", "limits.py"], "kind": "file", "value": "src/shiftdesk/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["src/shiftdesk/store.py", "store.py"], "kind": "file", "value": "src/shiftdesk/store.py"}, {"...
[{"answers": ["src/shiftdesk/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mo...
ctx_compress-s1-01624-twin
shiftdesk
ctx_compress-s1-01624
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] split_window double counts carried amounts when the period closes on a weekend. It lives in src/shiftdesk/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shiftdesk ...
39,705
1,985
63
108
false
synth:4476294910221231004@1.00388
synth
[{"aliases": ["src/shiftdesk/limits.py", "limits.py"], "kind": "file", "value": "src/shiftdesk/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["src/shiftdesk/store.py", "store.py"], "kind": "file", "value": "src/shiftdesk/store.py"}, {"...
[{"answers": ["src/shiftdesk/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mo...
ctx_compress-s1-01625
sectionledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] collate_draft leaves a trailing dash when the title ends in punctuation. It lives in src/sectionledger/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionledger l...
68,174
3,072
97
120
false
synth:5994769703333660347@1.22073
synth
[{"aliases": ["src/sectionledger/rules.py", "rules.py"], "kind": "file", "value": "src/sectionledger/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/sectionledger/limits.py", "limits.py"], "kind": "file", "value": "src/sectionledger/li...
[{"answers": ["src/sectionledger/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's ...
ctx_compress-s1-01625-twin
sectionledger
ctx_compress-s1-01625
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] collate_draft leaves a trailing dash when the title ends in punctuation. It lives in src/sectionledger/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionledger l...
33,097
1,654
55
120
false
synth:5994769703333660347@1.22073
synth
[{"aliases": ["src/sectionledger/rules.py", "rules.py"], "kind": "file", "value": "src/sectionledger/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/sectionledger/limits.py", "limits.py"], "kind": "file", "value": "src/sectionledger/li...
[{"answers": ["src/sectionledger/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's ...
ctx_compress-s1-01626
grantkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] revoke_grant keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/grantkit/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the grantkit layout. [to...
47,888
2,394
76
95
false
synth:6850135178789651218@1.12617
synth
[{"aliases": ["src/grantkit/cli.py", "cli.py"], "kind": "file", "value": "src/grantkit/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/grantkit/schema.py", "schema.py"], "kind": "file", "value": "src/grantkit/schema.py"}, {"aliases": ["src/gra...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["data/session_2.csv", "session_2.csv"]...
ctx_compress-s1-01626-twin
grantkit
ctx_compress-s1-01626
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] revoke_grant keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/grantkit/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the grantkit layout. [to...
26,811
1,340
43
95
false
synth:6850135178789651218@1.12617
synth
[{"aliases": ["src/grantkit/cli.py", "cli.py"], "kind": "file", "value": "src/grantkit/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/grantkit/schema.py", "schema.py"], "kind": "file", "value": "src/grantkit/schema.py"}, {"aliases": ["src/gra...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_rules.py", "tests/test_rules.py"...
ctx_compress-s1-01627
receiptforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] net_invoice raises RefundShapeError for bytes input in strict mode instead of decoding it. It lives in src/receiptforge/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
82,968
3,072
107
126
false
synth:1235922088205236904@1.77782
synth
[{"aliases": ["src/receiptforge/limits.py", "limits.py"], "kind": "file", "value": "src/receiptforge/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/tariff_7.csv", "tariff_7.csv"], "kind": "file", "value": "data/tariff_7.csv"}, {"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["credit_5.csv", "data/credit_5.csv"], ...
ctx_compress-s1-01627-twin
receiptforge
ctx_compress-s1-01627
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] net_invoice raises RefundShapeError for bytes input in strict mode instead of decoding it. It lives in src/receiptforge/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
42,821
2,141
57
126
false
synth:1235922088205236904@1.77782
synth
[{"aliases": ["src/receiptforge/limits.py", "limits.py"], "kind": "file", "value": "src/receiptforge/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/tariff_7.csv", "tariff_7.csv"], "kind": "file", "value": "data/tariff_7.csv"}, {"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["schema.py", "src/receiptforge/schema....
ctx_compress-s1-01628
tariffdesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] allocate_ledger raises KeyError for a missing key instead of returning the default. It lives in src/tariffdesk/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tarif...
104,534
3,072
118
139
false
synth:7891182916680209111@1.8631
synth
[{"aliases": ["src/tariffdesk/compat.py", "compat.py"], "kind": "file", "value": "src/tariffdesk/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"alia...
[{"answers": ["src/tariffdesk/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01628-twin
tariffdesk
ctx_compress-s1-01628
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] allocate_ledger raises KeyError for a missing key instead of returning the default. It lives in src/tariffdesk/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tarif...
63,301
3,072
72
139
false
synth:7891182916680209111@1.8631
synth
[{"aliases": ["src/tariffdesk/compat.py", "compat.py"], "kind": "file", "value": "src/tariffdesk/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"alia...
[{"answers": ["src/tariffdesk/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01629
revisionforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] paginate_margin keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/revisionforge/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the revisionforge ...
28,316
1,415
55
66
false
synth:2413001659819197460@1
synth
[{"aliases": ["src/revisionforge/io.py", "io.py"], "kind": "file", "value": "src/revisionforge/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/revisionforge/compat.py", "compat.py"], "kind": "file", "value": "src/revisionforge/compat.py"}, {"alias...
[{"answers": ["src/revisionforge/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most r...
ctx_compress-s1-01629-twin
revisionforge
ctx_compress-s1-01629
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] paginate_margin keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/revisionforge/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the revisionforge ...
18,285
1,000
39
66
false
synth:2413001659819197460@1
synth
[{"aliases": ["src/revisionforge/io.py", "io.py"], "kind": "file", "value": "src/revisionforge/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/revisionforge/compat.py", "compat.py"], "kind": "file", "value": "src/revisionforge/compat.py"}, {"alias...
[{"answers": ["src/revisionforge/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most r...
ctx_compress-s1-01630
sectionkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_footnote drops override values of 0, False and empty string. It lives in src/sectionkit/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionkit layout. [too...
40,586
2,029
74
87
false
synth:2843323939799501220@1
synth
[{"aliases": ["src/sectionkit/store.py", "store.py"], "kind": "file", "value": "src/sectionkit/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["src/sectionkit/core.py", "core.py"], "kind": "file", "value": "src/sectionkit/core.py"}, {"alias...
[{"answers": ["src/sectionkit/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mos...
ctx_compress-s1-01630-twin
sectionkit
ctx_compress-s1-01630
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] redact_footnote drops override values of 0, False and empty string. It lives in src/sectionkit/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionkit layout. [too...
26,562
1,328
48
87
false
synth:2843323939799501220@1
synth
[{"aliases": ["src/sectionkit/store.py", "store.py"], "kind": "file", "value": "src/sectionkit/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["src/sectionkit/core.py", "core.py"], "kind": "file", "value": "src/sectionkit/core.py"}, {"alias...
[{"answers": ["src/sectionkit/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mos...
ctx_compress-s1-01631
stemdesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] normalise_glyph raises ChunkBoundError for bytes input in strict mode instead of decoding it. It lives in src/stemdesk/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
93,684
3,072
113
130
false
synth:6183948180649613587@2.16855
synth
[{"aliases": ["src/stemdesk/limits.py", "limits.py"], "kind": "file", "value": "src/stemdesk/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases"...
[{"answers": ["src/stemdesk/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mos...
ctx_compress-s1-01631-twin
stemdesk
ctx_compress-s1-01631
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] normalise_glyph raises ChunkBoundError for bytes input in strict mode instead of decoding it. It lives in src/stemdesk/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
47,285
2,364
59
130
false
synth:6183948180649613587@2.16855
synth
[{"aliases": ["src/stemdesk/limits.py", "limits.py"], "kind": "file", "value": "src/stemdesk/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["cli.py", "src/stemdesk/cli.py"], "kin...
ctx_compress-s1-01632
invoiceflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reconcile_payout averages one value too many once the window is full. It lives in src/invoiceflow/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the invoiceflow layout. ...
33,352
1,667
56
63
false
synth:3116438993786578792@1
synth
[{"aliases": ["src/invoiceflow/store.py", "store.py"], "kind": "file", "value": "src/invoiceflow/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["data/ledger_4.csv", "ledger_4.csv"], "kind": "file", "value": "data/ledger_4.csv"}, {"aliases"...
[{"answers": ["src/invoiceflow/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mo...
ctx_compress-s1-01632-twin
invoiceflow
ctx_compress-s1-01632
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reconcile_payout averages one value too many once the window is full. It lives in src/invoiceflow/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the invoiceflow layout. ...
18,039
1,000
34
63
false
synth:3116438993786578792@1
synth
[{"aliases": ["src/invoiceflow/store.py", "store.py"], "kind": "file", "value": "src/invoiceflow/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["data/ledger_4.csv", "ledger_4.csv"], "kind": "file", "value": "data/ledger_4.csv"}, {"aliases"...
[{"answers": ["src/invoiceflow/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mo...
ctx_compress-s1-01633
tariffflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] accrue_refund returns the upper bound for every value instead of clamping into the range. It lives in src/tariffflow/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the t...
35,157
1,757
64
79
false
synth:2142818404594063151@1
synth
[{"aliases": ["src/tariffflow/store.py", "store.py"], "kind": "file", "value": "src/tariffflow/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["data/balance_3.csv", "balance_3.csv"], "kind": "file", "value": "data/balance_3.csv"}, {"aliases...
[{"answers": ["src/tariffflow/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mos...
ctx_compress-s1-01633-twin
tariffflow
ctx_compress-s1-01633
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] accrue_refund returns the upper bound for every value instead of clamping into the range. It lives in src/tariffflow/store.py; `python -m pytest tests/test_store.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the t...
17,155
1,000
38
79
false
synth:2142818404594063151@1
synth
[{"aliases": ["src/tariffflow/store.py", "store.py"], "kind": "file", "value": "src/tariffflow/store.py"}, {"aliases": ["tests/test_store.py", "test_store.py"], "kind": "file", "value": "tests/test_store.py"}, {"aliases": ["data/balance_3.csv", "balance_3.csv"], "kind": "file", "value": "data/balance_3.csv"}, {"aliases...
[{"answers": ["src/tariffflow/store.py", "store.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mos...
ctx_compress-s1-01634
zoneforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bound_bearing_once produces doubled dashes for consecutive separators. It lives in src/zoneforge/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the zoneforge layout. [tool_u...
54,761
2,738
104
119
false
synth:454071860011634563@1.1244
synth
[{"aliases": ["src/zoneforge/cli.py", "cli.py"], "kind": "file", "value": "src/zoneforge/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases": ["data/corrido...
[{"answers": ["src/zoneforge/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most rec...
ctx_compress-s1-01634-twin
zoneforge
ctx_compress-s1-01634
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bound_bearing_once produces doubled dashes for consecutive separators. It lives in src/zoneforge/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the zoneforge layout. [tool_u...
26,421
1,321
54
119
false
synth:454071860011634563@1.1244
synth
[{"aliases": ["src/zoneforge/cli.py", "cli.py"], "kind": "file", "value": "src/zoneforge/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["docs/changelog.md", "changelog.md"], "kind": "file", "value": "docs/CHANGELOG.md"}, {"aliases": ["data/corrido...
[{"answers": ["src/zoneforge/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most rec...
ctx_compress-s1-01635
sectionflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] anchor_draft ignores the cap, so late attempts wait for minutes. It lives in src/sectionflow/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionflow layout. [tool...
75,620
3,072
97
124
false
synth:1663308476970221294@2.1913
synth
[{"aliases": ["src/sectionflow/rules.py", "rules.py"], "kind": "file", "value": "src/sectionflow/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/sectionflow/audit.py", "audit.py"], "kind": "file", "value": "src/sectionflow/audit.py"}, ...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["data/index_0.csv", "index_0.csv"], "k...
ctx_compress-s1-01635-twin
sectionflow
ctx_compress-s1-01635
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] anchor_draft ignores the cap, so late attempts wait for minutes. It lives in src/sectionflow/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the sectionflow layout. [tool...
49,393
2,469
71
124
false
synth:1663308476970221294@2.1913
synth
[{"aliases": ["src/sectionflow/rules.py", "rules.py"], "kind": "file", "value": "src/sectionflow/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/sectionflow/audit.py", "audit.py"], "kind": "file", "value": "src/sectionflow/audit.py"}, ...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["data/heading_3.csv", "heading_3.csv"]...
ctx_compress-s1-01636
ledgerworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] net_balance lets defaults overwrite explicit overrides. It lives in src/ledgerworks/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the ledgerworks layout. [tool_use kind=searc...
46,027
2,301
79
90
false
synth:1957708900754507465@1
synth
[{"aliases": ["src/ledgerworks/io.py", "io.py"], "kind": "file", "value": "src/ledgerworks/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/ledgerworks/report.py", "report.py"], "kind": "file", "value": "src/ledgerworks/report.py"}, {"aliases": ["s...
[{"answers": ["src/ledgerworks/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most rec...
ctx_compress-s1-01636-twin
ledgerworks
ctx_compress-s1-01636
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] net_balance lets defaults overwrite explicit overrides. It lives in src/ledgerworks/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the ledgerworks layout. [tool_use kind=searc...
21,633
1,081
43
90
false
synth:1957708900754507465@1
synth
[{"aliases": ["src/ledgerworks/io.py", "io.py"], "kind": "file", "value": "src/ledgerworks/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/ledgerworks/report.py", "report.py"], "kind": "file", "value": "src/ledgerworks/report.py"}, {"aliases": ["s...
[{"answers": ["src/ledgerworks/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most rec...
ctx_compress-s1-01637
seriesworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_counter skips the first page; page 1 returns what should be page 2. It lives in src/seriesworks/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the seriesworks layout. ...
90,081
3,072
105
118
false
synth:4358817905453023072@2.37308
synth
[{"aliases": ["src/seriesworks/core.py", "core.py"], "kind": "file", "value": "src/seriesworks/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["data/bucket_4.csv", "bucket_4.csv"], "kind": "file", "value": "data/bucket_4.csv"}, {"aliases": ["sr...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["counter_3.csv", "data/counter_3.csv"]...
ctx_compress-s1-01637-twin
seriesworks
ctx_compress-s1-01637
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_counter skips the first page; page 1 returns what should be page 2. It lives in src/seriesworks/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the seriesworks layout. ...
44,792
2,239
59
118
false
synth:4358817905453023072@2.37308
synth
[{"aliases": ["src/seriesworks/core.py", "core.py"], "kind": "file", "value": "src/seriesworks/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["data/bucket_4.csv", "bucket_4.csv"], "kind": "file", "value": "data/bucket_4.csv"}, {"aliases": ["sr...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["counter_3.csv", "data/counter_3.csv"]...
ctx_compress-s1-01638
palletflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_lot_batch raises WarehouseShapeError for bytes input in strict mode instead of decoding it. It lives in src/palletflow/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
29,983
1,499
57
68
false
synth:1723189563072239592@1
synth
[{"aliases": ["src/palletflow/core.py", "core.py"], "kind": "file", "value": "src/palletflow/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/palletflow/schema.py", "schema.py"], "kind": "file", "value": "src/palletflow/schema.py"}, {"alias...
[{"answers": ["src/palletflow/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01638-twin
palletflow
ctx_compress-s1-01638
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_lot_batch raises WarehouseShapeError for bytes input in strict mode instead of decoding it. It lives in src/palletflow/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping t...
13,355
1,000
33
68
false
synth:1723189563072239592@1
synth
[{"aliases": ["src/palletflow/core.py", "core.py"], "kind": "file", "value": "src/palletflow/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/palletflow/schema.py", "schema.py"], "kind": "file", "value": "src/palletflow/schema.py"}, {"alias...
[{"answers": ["src/palletflow/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most ...
ctx_compress-s1-01639
lotflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_stock_all raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/lotflow/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the lotflo...
37,216
1,860
80
99
false
synth:1733897040153305779@1.2849
synth
[{"aliases": ["src/lotflow/compat.py", "compat.py"], "kind": "file", "value": "src/lotflow/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/lotflow/report.py", "report.py"], "kind": "file", "value": "src/lotflow/report.py"}, {"alias...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["changelog.md", "docs/changelog.md"], ...
ctx_compress-s1-01639-twin
lotflow
ctx_compress-s1-01639
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] age_stock_all raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/lotflow/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the lotflo...
24,311
1,215
57
99
false
synth:1733897040153305779@1.2849
synth
[{"aliases": ["src/lotflow/compat.py", "compat.py"], "kind": "file", "value": "src/lotflow/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/lotflow/report.py", "report.py"], "kind": "file", "value": "src/lotflow/report.py"}, {"alias...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["schema.py", "src/lotflow/schema.py"],...
ctx_compress-s1-01640
fixkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_zone_all raises ZoneShapeError for bytes input in strict mode instead of decoding it. It lives in src/fixkit/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the fixki...
74,840
3,072
99
126
false
synth:5093027332714794418@1.5754
synth
[{"aliases": ["src/fixkit/cli.py", "cli.py"], "kind": "file", "value": "src/fixkit/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/fix_7.csv", "fix_7.csv"], "kind": "file", "value": "data/fix_7.csv"}, {"aliases": ["data/tile_6.csv", "tile_6.c...
[{"answers": ["src/fixkit/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent...
ctx_compress-s1-01640-twin
fixkit
ctx_compress-s1-01640
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_zone_all raises ZoneShapeError for bytes input in strict mode instead of decoding it. It lives in src/fixkit/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the fixki...
40,594
2,029
61
126
false
synth:5093027332714794418@1.5754
synth
[{"aliases": ["src/fixkit/cli.py", "cli.py"], "kind": "file", "value": "src/fixkit/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["data/fix_7.csv", "fix_7.csv"], "kind": "file", "value": "data/fix_7.csv"}, {"aliases": ["data/tile_6.csv", "tile_6.c...
[{"answers": ["src/fixkit/cli.py", "cli.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent...
ctx_compress-s1-01641
tracekit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] clip_bucket raises ValueError when a value itself contains an equals sign. It lives in src/tracekit/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tracekit layout. [to...
61,376
3,068
98
111
false
synth:2500781882961555620@1.05088
synth
[{"aliases": ["src/tracekit/core.py", "core.py"], "kind": "file", "value": "src/tracekit/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/tracekit/audit.py", "audit.py"], "kind": "file", "value": "src/tracekit/audit.py"}, {"aliases": ["data...
[{"answers": ["src/tracekit/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most re...
ctx_compress-s1-01641-twin
tracekit
ctx_compress-s1-01641
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] clip_bucket raises ValueError when a value itself contains an equals sign. It lives in src/tracekit/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the tracekit layout. [to...
34,407
1,720
50
111
false
synth:2500781882961555620@1.05088
synth
[{"aliases": ["src/tracekit/core.py", "core.py"], "kind": "file", "value": "src/tracekit/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/tracekit/audit.py", "audit.py"], "kind": "file", "value": "src/tracekit/audit.py"}, {"aliases": ["data...
[{"answers": ["src/tracekit/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most re...
ctx_compress-s1-01642
refundforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] roll_tariff double counts carried amounts when the period closes on a weekend. It lives in src/refundforge/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the refundforge l...
84,446
3,072
95
102
false
synth:8476046808924891550@1.60082
synth
[{"aliases": ["src/refundforge/core.py", "core.py"], "kind": "file", "value": "src/refundforge/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/refundforge/report.py", "report.py"], "kind": "file", "value": "src/refundforge/report.py"}, {"a...
[{"answers": ["src/refundforge/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most...
ctx_compress-s1-01642-twin
refundforge
ctx_compress-s1-01642
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] roll_tariff double counts carried amounts when the period closes on a weekend. It lives in src/refundforge/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the refundforge l...
41,925
2,096
51
102
false
synth:8476046808924891550@1.60082
synth
[{"aliases": ["src/refundforge/core.py", "core.py"], "kind": "file", "value": "src/refundforge/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/refundforge/report.py", "report.py"], "kind": "file", "value": "src/refundforge/report.py"}, {"a...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_store.py", "tests/test_store.py"...
ctx_compress-s1-01643
balanceworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] roll_credit_window drops override values of 0, False and empty string. It lives in src/balanceworks/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the balanceworks layou...
20,193
1,009
48
61
false
synth:5115212933566429207@1
synth
[{"aliases": ["src/balanceworks/rules.py", "rules.py"], "kind": "file", "value": "src/balanceworks/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/balanceworks/core.py", "core.py"], "kind": "file", "value": "src/balanceworks/core.py"},...
[{"answers": ["src/balanceworks/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's m...
ctx_compress-s1-01643-twin
balanceworks
ctx_compress-s1-01643
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] roll_credit_window drops override values of 0, False and empty string. It lives in src/balanceworks/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the balanceworks layou...
12,743
1,000
35
61
false
synth:5115212933566429207@1
synth
[{"aliases": ["src/balanceworks/rules.py", "rules.py"], "kind": "file", "value": "src/balanceworks/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["src/balanceworks/core.py", "core.py"], "kind": "file", "value": "src/balanceworks/core.py"},...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["rules.py", "src/balanceworks/rules.py...
ctx_compress-s1-01644
skuforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_shipment_once raises ValueError when a value itself contains an equals sign. It lives in src/skuforge/format.py; `python -m pytest tests/test_format.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the skufor...
63,826
3,072
96
113
false
synth:2976899231616086874@1.4871
synth
[{"aliases": ["src/skuforge/format.py", "format.py"], "kind": "file", "value": "src/skuforge/format.py"}, {"aliases": ["tests/test_format.py", "test_format.py"], "kind": "file", "value": "tests/test_format.py"}, {"aliases": ["data/stock_1.csv", "stock_1.csv"], "kind": "file", "value": "data/stock_1.csv"}, {"aliases": [...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_rules.py", "tests/test_rules.py"...
ctx_compress-s1-01644-twin
skuforge
ctx_compress-s1-01644
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_shipment_once raises ValueError when a value itself contains an equals sign. It lives in src/skuforge/format.py; `python -m pytest tests/test_format.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the skufor...
36,265
1,813
59
113
false
synth:2976899231616086874@1.4871
synth
[{"aliases": ["src/skuforge/format.py", "format.py"], "kind": "file", "value": "src/skuforge/format.py"}, {"aliases": ["tests/test_format.py", "test_format.py"], "kind": "file", "value": "tests/test_format.py"}, {"aliases": ["data/stock_1.csv", "stock_1.csv"], "kind": "file", "value": "data/stock_1.csv"}, {"aliases": [...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["compat.py", "src/skuforge/compat.py"]...
ctx_compress-s1-01645
shipmentledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_warehouse produces doubled dashes for consecutive separators. It lives in src/shipmentledger/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shipmentledger layout. ...
37,351
1,867
79
88
false
synth:6709259704623124730@1
synth
[{"aliases": ["src/shipmentledger/io.py", "io.py"], "kind": "file", "value": "src/shipmentledger/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/shipmentledger/compat.py", "compat.py"], "kind": "file", "value": "src/shipmentledger/compat.py"}, {"a...
[{"answers": ["src/shipmentledger/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01645-twin
shipmentledger
ctx_compress-s1-01645
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] reserve_warehouse produces doubled dashes for consecutive separators. It lives in src/shipmentledger/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the shipmentledger layout. ...
23,786
1,189
49
88
false
synth:6709259704623124730@1
synth
[{"aliases": ["src/shipmentledger/io.py", "io.py"], "kind": "file", "value": "src/shipmentledger/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/shipmentledger/compat.py", "compat.py"], "kind": "file", "value": "src/shipmentledger/compat.py"}, {"a...
[{"answers": ["src/shipmentledger/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most ...
ctx_compress-s1-01646
cueworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] hyphenate_snippet_all lets defaults overwrite explicit overrides. It lives in src/cueworks/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the cueworks layout. [tool_use kind...
58,001
2,900
113
126
false
synth:2135760398740353107@1.04757
synth
[{"aliases": ["src/cueworks/cli.py", "cli.py"], "kind": "file", "value": "src/cueworks/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/cueworks/schema.py", "schema.py"], "kind": "file", "value": "src/cueworks/schema.py"}, {"aliases": ["src/cue...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["src/cueworks/tables.py", "tables.py"]...
ctx_compress-s1-01646-twin
cueworks
ctx_compress-s1-01646
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] hyphenate_snippet_all lets defaults overwrite explicit overrides. It lives in src/cueworks/cli.py; `python -m pytest tests/test_cli.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the cueworks layout. [tool_use kind...
34,254
1,712
63
126
false
synth:2135760398740353107@1.04757
synth
[{"aliases": ["src/cueworks/cli.py", "cli.py"], "kind": "file", "value": "src/cueworks/cli.py"}, {"aliases": ["tests/test_cli.py", "test_cli.py"], "kind": "file", "value": "tests/test_cli.py"}, {"aliases": ["src/cueworks/schema.py", "schema.py"], "kind": "file", "value": "src/cueworks/schema.py"}, {"aliases": ["src/cue...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["chunk_0.csv", "data/chunk_0.csv"], "k...
ctx_compress-s1-01647
sessiondesk
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] verify_nonce raises SessionShapeError for bytes input in strict mode instead of decoding it. It lives in src/sessiondesk/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the ses...
57,364
2,868
96
121
false
synth:2609911059674103928@1.25395
synth
[{"aliases": ["src/sessiondesk/io.py", "io.py"], "kind": "file", "value": "src/sessiondesk/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/grant_2.csv", "grant_2.csv"], "kind": "file", "value": "data/grant_2.csv"}, {"aliases": ["src/sessiondesk/f...
[{"answers": ["src/sessiondesk/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most rec...
ctx_compress-s1-01647-twin
sessiondesk
ctx_compress-s1-01647
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] verify_nonce raises SessionShapeError for bytes input in strict mode instead of decoding it. It lives in src/sessiondesk/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the ses...
32,442
1,622
68
121
false
synth:2609911059674103928@1.25395
synth
[{"aliases": ["src/sessiondesk/io.py", "io.py"], "kind": "file", "value": "src/sessiondesk/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/grant_2.csv", "grant_2.csv"], "kind": "file", "value": "data/grant_2.csv"}, {"aliases": ["src/sessiondesk/f...
[{"answers": ["src/sessiondesk/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most rec...
ctx_compress-s1-01648
scopeworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] revoke_session_all grows linearly instead of doubling. It lives in src/scopeworks/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the scopeworks layout. [tool_use kind=...
79,134
3,072
99
110
false
synth:735652353438528037@1.7354
synth
[{"aliases": ["src/scopeworks/report.py", "report.py"], "kind": "file", "value": "src/scopeworks/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/scopeworks/schema.py", "schema.py"], "kind": "file", "value": "src/scopeworks/schema.p...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["report.py", "src/scopeworks/report.py...
ctx_compress-s1-01648-twin
scopeworks
ctx_compress-s1-01648
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] revoke_session_all grows linearly instead of doubling. It lives in src/scopeworks/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the scopeworks layout. [tool_use kind=...
43,950
2,197
63
110
false
synth:735652353438528037@1.7354
synth
[{"aliases": ["src/scopeworks/report.py", "report.py"], "kind": "file", "value": "src/scopeworks/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/scopeworks/schema.py", "schema.py"], "kind": "file", "value": "src/scopeworks/schema.p...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["changelog.md", "docs/changelog.md"], ...
ctx_compress-s1-01649
bucketkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_trace_strict raises KeyError for a missing key instead of returning the default. It lives in src/bucketkit/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the bucke...
78,495
3,072
114
125
false
synth:60570568115999424@1.69955
synth
[{"aliases": ["src/bucketkit/compat.py", "compat.py"], "kind": "file", "value": "src/bucketkit/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/bucketkit/limits.py", "limits.py"], "kind": "file", "value": "src/bucketkit/limits.py"},...
[{"answers": ["src/bucketkit/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mo...
ctx_compress-s1-01649-twin
bucketkit
ctx_compress-s1-01649
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_trace_strict raises KeyError for a missing key instead of returning the default. It lives in src/bucketkit/compat.py; `python -m pytest tests/test_compat.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the bucke...
42,945
2,147
70
125
false
synth:60570568115999424@1.69955
synth
[{"aliases": ["src/bucketkit/compat.py", "compat.py"], "kind": "file", "value": "src/bucketkit/compat.py"}, {"aliases": ["tests/test_compat.py", "test_compat.py"], "kind": "file", "value": "tests/test_compat.py"}, {"aliases": ["src/bucketkit/limits.py", "limits.py"], "kind": "file", "value": "src/bucketkit/limits.py"},...
[{"answers": ["src/bucketkit/compat.py", "compat.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's mo...
ctx_compress-s1-01650
principalflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] extend_session skips the first page; page 1 returns what should be page 2. It lives in src/principalflow/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the principalfl...
94,959
3,072
104
123
false
synth:2918176474444622860@1.86042
synth
[{"aliases": ["src/principalflow/limits.py", "limits.py"], "kind": "file", "value": "src/principalflow/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/scope_4.csv", "scope_4.csv"], "kind": "file", "value": "data/scope_4.csv"}, {"a...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["data/session_1.csv", "session_1.csv"]...
ctx_compress-s1-01650-twin
principalflow
ctx_compress-s1-01650
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] extend_session skips the first page; page 1 returns what should be page 2. It lives in src/principalflow/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the principalfl...
62,754
3,072
70
123
false
synth:2918176474444622860@1.86042
synth
[{"aliases": ["src/principalflow/limits.py", "limits.py"], "kind": "file", "value": "src/principalflow/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["data/scope_4.csv", "scope_4.csv"], "kind": "file", "value": "data/scope_4.csv"}, {"a...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["io.py", "src/principalflow/io.py"], "...
ctx_compress-s1-01651
quotaforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] extend_slot returns one item too few on every page. It lives in src/quotaforge/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the quotaforge layout. [tool_use kind=search name...
68,166
3,072
106
123
false
synth:319597856883965962@1.6126
synth
[{"aliases": ["src/quotaforge/io.py", "io.py"], "kind": "file", "value": "src/quotaforge/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/quotaforge/core.py", "core.py"], "kind": "file", "value": "src/quotaforge/core.py"}, {"aliases": ["src/quotafo...
[{"answers": ["src/quotaforge/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most rece...
ctx_compress-s1-01651-twin
quotaforge
ctx_compress-s1-01651
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] extend_slot returns one item too few on every page. It lives in src/quotaforge/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the quotaforge layout. [tool_use kind=search name...
27,156
1,357
56
123
false
synth:319597856883965962@1.6126
synth
[{"aliases": ["src/quotaforge/io.py", "io.py"], "kind": "file", "value": "src/quotaforge/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["src/quotaforge/core.py", "core.py"], "kind": "file", "value": "src/quotaforge/core.py"}, {"aliases": ["src/quotafo...
[{"answers": ["src/quotaforge/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most rece...
ctx_compress-s1-01652
gaugeledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_trace lets defaults overwrite explicit overrides. It lives in src/gaugeledger/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the gaugeledger layout. [tool_use kind...
45,095
2,254
85
104
false
synth:7710183509703911558@1
synth
[{"aliases": ["src/gaugeledger/report.py", "report.py"], "kind": "file", "value": "src/gaugeledger/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["data/sample_2.csv", "sample_2.csv"], "kind": "file", "value": "data/sample_2.csv"}, {"al...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["test_format.py", "tests/test_format.p...
ctx_compress-s1-01652-twin
gaugeledger
ctx_compress-s1-01652
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_trace lets defaults overwrite explicit overrides. It lives in src/gaugeledger/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the gaugeledger layout. [tool_use kind...
31,383
1,569
61
104
false
synth:7710183509703911558@1
synth
[{"aliases": ["src/gaugeledger/report.py", "report.py"], "kind": "file", "value": "src/gaugeledger/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["data/sample_2.csv", "sample_2.csv"], "kind": "file", "value": "data/sample_2.csv"}, {"al...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["core.py", "src/gaugeledger/core.py"],...
ctx_compress-s1-01653
gaugeflow
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_counter keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/gaugeflow/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the gaugeflow layout. ...
41,639
2,081
76
81
false
synth:5366087963592095900@1
synth
[{"aliases": ["src/gaugeflow/core.py", "core.py"], "kind": "file", "value": "src/gaugeflow/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["data/sample_4.csv", "sample_4.csv"], "kind": "file", "value": "data/sample_4.csv"}, {"aliases": ["src/ga...
[{"answers": ["src/gaugeflow/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's most r...
ctx_compress-s1-01653-twin
gaugeflow
ctx_compress-s1-01653
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] bin_counter keeps leading spaces in keys so lookups after a semicolon fail. It lives in src/gaugeflow/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the gaugeflow layout. ...
32,584
1,629
48
81
false
synth:5366087963592095900@1
synth
[{"aliases": ["src/gaugeflow/core.py", "core.py"], "kind": "file", "value": "src/gaugeflow/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["data/sample_4.csv", "sample_4.csv"], "kind": "file", "value": "data/sample_4.csv"}, {"aliases": ["src/ga...
[{"answers": ["src/gaugeflow/core.py", "core.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most r...
ctx_compress-s1-01654
cueworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] normalise_stem returns one item too few on every page. It lives in src/cueworks/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the cueworks layout. [tool_use kind=search name=...
110,542
3,072
109
124
false
synth:3426592043650636954@1.85478
synth
[{"aliases": ["src/cueworks/io.py", "io.py"], "kind": "file", "value": "src/cueworks/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/stem_0.csv", "stem_0.csv"], "kind": "file", "value": "data/stem_0.csv"}, {"aliases": ["data/snippet_8.csv", "snip...
[{"answers": ["src/cueworks/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent...
ctx_compress-s1-01654-twin
cueworks
ctx_compress-s1-01654
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] normalise_stem returns one item too few on every page. It lives in src/cueworks/io.py; `python -m pytest tests/test_io.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the cueworks layout. [tool_use kind=search name=...
52,505
2,625
61
124
false
synth:3426592043650636954@1.85478
synth
[{"aliases": ["src/cueworks/io.py", "io.py"], "kind": "file", "value": "src/cueworks/io.py"}, {"aliases": ["tests/test_io.py", "test_io.py"], "kind": "file", "value": "tests/test_io.py"}, {"aliases": ["data/stem_0.csv", "stem_0.csv"], "kind": "file", "value": "data/stem_0.csv"}, {"aliases": ["data/snippet_8.csv", "snip...
[{"answers": ["src/cueworks/io.py", "io.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent...
ctx_compress-s1-01655
rosterforge
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] split_booking raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/rosterforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the rost...
83,956
3,072
88
97
false
synth:1841396123586935879@1.75127
synth
[{"aliases": ["src/rosterforge/rules.py", "rules.py"], "kind": "file", "value": "src/rosterforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["data/quota_10.csv", "quota_10.csv"], "kind": "file", "value": "data/quota_10.csv"}, {"aliases"...
[{"answers": ["src/rosterforge/rules.py", "rules.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's mo...
ctx_compress-s1-01655-twin
rosterforge
ctx_compress-s1-01655
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] split_booking raises ZeroDivisionError on the first value and lags by one afterwards. It lives in src/rosterforge/rules.py; `python -m pytest tests/test_rules.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the rost...
43,928
2,196
48
97
false
synth:1841396123586935879@1.75127
synth
[{"aliases": ["src/rosterforge/rules.py", "rules.py"], "kind": "file", "value": "src/rosterforge/rules.py"}, {"aliases": ["tests/test_rules.py", "test_rules.py"], "kind": "file", "value": "tests/test_rules.py"}, {"aliases": ["data/quota_10.csv", "quota_10.csv"], "kind": "file", "value": "data/quota_10.csv"}, {"aliases"...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["calendar_6.csv", "data/calendar_6.csv...
ctx_compress-s1-01656
revisionkit
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] anchor_footnote returns the upper bound for every value instead of clamping into the range. It lives in src/revisionkit/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
36,451
1,822
76
81
false
synth:6527560983784575226@1
synth
[{"aliases": ["src/revisionkit/report.py", "report.py"], "kind": "file", "value": "src/revisionkit/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/revisionkit/compat.py", "compat.py"], "kind": "file", "value": "src/revisionkit/comp...
[{"answers": ["src/revisionkit/report.py", "report.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["fail"], "question": "Did the agent's ...
ctx_compress-s1-01656-twin
revisionkit
ctx_compress-s1-01656
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] anchor_footnote returns the upper bound for every value instead of clamping into the range. It lives in src/revisionkit/report.py; `python -m pytest tests/test_report.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping ...
27,897
1,394
46
81
false
synth:6527560983784575226@1
synth
[{"aliases": ["src/revisionkit/report.py", "report.py"], "kind": "file", "value": "src/revisionkit/report.py"}, {"aliases": ["tests/test_report.py", "test_report.py"], "kind": "file", "value": "tests/test_report.py"}, {"aliases": ["src/revisionkit/compat.py", "compat.py"], "kind": "file", "value": "src/revisionkit/comp...
[{"answers": ["src/revisionkit/report.py", "report.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's ...
ctx_compress-s1-01657
corridorledger
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_beacon grows linearly instead of doubling. It lives in src/corridorledger/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the corridorledger layout. [tool_use kind=...
50,672
2,533
82
99
false
synth:7879385119545365174@1.375
synth
[{"aliases": ["src/corridorledger/core.py", "core.py"], "kind": "file", "value": "src/corridorledger/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/corridorledger/limits.py", "limits.py"], "kind": "file", "value": "src/corridorledger/limi...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["beacon_1.csv", "data/beacon_1.csv"], ...
ctx_compress-s1-01657-twin
corridorledger
ctx_compress-s1-01657
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] cluster_beacon grows linearly instead of doubling. It lives in src/corridorledger/core.py; `python -m pytest tests/test_core.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by mapping the corridorledger layout. [tool_use kind=...
30,664
1,533
49
99
false
synth:7879385119545365174@1.375
synth
[{"aliases": ["src/corridorledger/core.py", "core.py"], "kind": "file", "value": "src/corridorledger/core.py"}, {"aliases": ["tests/test_core.py", "test_core.py"], "kind": "file", "value": "tests/test_core.py"}, {"aliases": ["src/corridorledger/limits.py", "limits.py"], "kind": "file", "value": "src/corridorledger/limi...
[{"answers": ["no"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's most recent test run pass or fail? Answer pass or fail.", "type": "word"}, {"answers": {"file": ["limits.py", "src/corridorledger/limit...
ctx_compress-s1-01658
bearingworks
You compact a coding agent's session so the agent can continue from your summary alone. The session is inside <session> tags: [user] messages, [assistant] text, [tool_use] calls with their kind and file, and [tool_result] outputs. Write the summary in plain markdown with exactly these nine headings, in this order, each...
<session> [user] interpolate_waypoint raises LegStateError for bytes input in strict mode instead of decoding it. It lives in src/bearingworks/limits.py; `python -m pytest tests/test_limits.py -q` shows the failures. Fix it without changing the public signature and add a regression test. [assistant] I will start by ma...
57,442
2,872
74
87
false
synth:119000594857403598@1
synth
[{"aliases": ["src/bearingworks/limits.py", "limits.py"], "kind": "file", "value": "src/bearingworks/limits.py"}, {"aliases": ["tests/test_limits.py", "test_limits.py"], "kind": "file", "value": "tests/test_limits.py"}, {"aliases": ["src/bearingworks/store.py", "store.py"], "kind": "file", "value": "src/bearingworks/st...
[{"answers": ["src/bearingworks/limits.py", "limits.py"], "question": "Which file did the agent edit most recently?", "type": "exact"}, {"answers": ["yes"], "question": "Has the agent changed any existing file in the repository yet? Answer yes or no.", "type": "word"}, {"answers": ["pass"], "question": "Did the agent's...