File size: 9,862 Bytes
83d107e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
{
  "@context": {
    "@language": "en",
    "@vocab": "https://schema.org/",
    "cr": "http://mlcommons.org/croissant/1.0/",
    "rai": "http://mlcommons.org/croissant/RAI/",
    "column": "cr:column",
    "dataType": { "@id": "cr:dataType", "@type": "@vocab" },
    "extract": "cr:extract",
    "field": "cr:field",
    "fileObject": "cr:fileObject",
    "format": "cr:format",
    "includes": "cr:includes",
    "isEnumeration": { "@id": "cr:isEnumeration", "@type": "sc:Boolean" },
    "jsonPath": "cr:jsonPath",
    "key": "cr:key",
    "md5": "cr:md5",
    "parentField": "cr:parentField",
    "path": "cr:path",
    "recordSet": "cr:recordSet",
    "references": "cr:references",
    "regex": "cr:regex",
    "repeated": { "@id": "cr:repeated", "@type": "sc:Boolean" },
    "separator": "cr:separator",
    "source": "cr:source",
    "subField": "cr:subField",
    "transform": "cr:transform",
    "sc": "https://schema.org/"
  },
  "@type": "Dataset",
  "@id": "https://huggingface.co/datasets/thakicloud/SkillRetBench",
  "name": "SkillRetBench",
  "description": "Query-to-skill retrieval benchmark: 501 procedural agent skills, 1,250 annotated queries across five settings (single_skill, multi_skill_composition, distractor, outdated_redundant, budget_constrained), and packaged baseline IR results (BM25, Dense, Hybrid, NaiveLLM, SADO). Languages: English and Korean. Companion to SEFO (Self-Evolving Federated Orchestration).",
  "url": "https://huggingface.co/datasets/thakicloud/SkillRetBench",
  "license": "https://www.apache.org/licenses/LICENSE-2.0",
  "creator": {
    "@type": "Organization",
    "name": "ThakiCloud AI Research"
  },
  "citation": "@misc{han2026sefo, title = {{SEFO}: Self-Evolving Federated Orchestration with Trusted Skill Governance and Skill Retrieval Benchmark for Recursive Agentic Systems}, author = {Han, Hyojung and {ThakiCloud AI Research}}, year = {2026}, howpublished = {\\url{https://github.com/hyojunguy/ai-model-event-stock-analytics}}, note = {Includes SkillRetBench}}",
  "version": "1.0.0",
  "distribution": [
    {
      "@type": "cr:FileObject",
      "@id": "skill_corpus.json",
      "name": "skill_corpus.json",
      "description": "Skill index with meta, statistics, and 501 skill records (ids, descriptions, triggers, full_text, token_count, composition edges).",
      "contentUrl": "https://huggingface.co/datasets/thakicloud/SkillRetBench/resolve/main/skill_corpus.json",
      "encodingFormat": "application/json",
      "sha256": ""
    },
    {
      "@type": "cr:FileObject",
      "@id": "skillretbench_queries.json",
      "name": "skillretbench_queries.json",
      "description": "1,250 benchmark queries with gold_skills, settings, difficulty, source, and corpus meta (rng_seed 42).",
      "contentUrl": "https://huggingface.co/datasets/thakicloud/SkillRetBench/resolve/main/skillretbench_queries.json",
      "encodingFormat": "application/json",
      "sha256": ""
    },
    {
      "@type": "cr:FileObject",
      "@id": "baseline_results.json",
      "name": "baseline_results.json",
      "description": "Per-baseline, per-setting IR metrics (Recall@k, nDCG@k, MRR, MAP) and summary_table markdown.",
      "contentUrl": "https://huggingface.co/datasets/thakicloud/SkillRetBench/resolve/main/baseline_results.json",
      "encodingFormat": "application/json",
      "sha256": ""
    }
  ],
  "recordSet": [
    {
      "@type": "cr:RecordSet",
      "@id": "recordset/skills",
      "name": "skills",
      "description": "One row per skill in the fixed retrieval corpus (aligned iteration over $.skills[*]).",
      "key": { "@id": "field/skill_id" },
      "source": {
        "fileObject": { "@id": "skill_corpus.json" },
        "extract": { "jsonPath": "$.skills[*]" }
      },
      "field": [
        {
          "@type": "cr:Field",
          "@id": "field/skill_id",
          "name": "skill_id",
          "description": "Canonical skill identifier (slug).",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.skill_id" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/skill_name",
          "name": "skill_name",
          "description": "Human-readable skill title.",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.skill_name" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/description",
          "name": "description",
          "description": "Short routing description (YAML description / summary).",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.description" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/category",
          "name": "category",
          "description": "Folder or taxonomy category for the skill.",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.category" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/token_count",
          "name": "token_count",
          "description": "Approximate token count of full skill body.",
          "dataType": "sc:Integer",
          "source": {
            "extract": { "jsonPath": "$.token_count" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/full_text",
          "name": "full_text",
          "description": "Full SKILL.md-derived markdown body used as retrieval document.",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.full_text" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/trigger_phrases",
          "name": "trigger_phrases",
          "description": "Quoted trigger phrases from skill spec.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "extract": { "jsonPath": "$.trigger_phrases[*]" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/korean_triggers",
          "name": "korean_triggers",
          "description": "Korean trigger tokens if present.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "extract": { "jsonPath": "$.korean_triggers[*]" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/composable_skills",
          "name": "composable_skills",
          "description": "Referenced skill ids for composition edges.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "extract": { "jsonPath": "$.composable_skills[*]" }
          }
        }
      ]
    },
    {
      "@type": "cr:RecordSet",
      "@id": "recordset/queries",
      "name": "queries",
      "description": "One row per benchmark query with gold and distractor skill lists (aligned iteration over $.queries[*]).",
      "key": { "@id": "field/query_id" },
      "source": {
        "fileObject": { "@id": "skillretbench_queries.json" },
        "extract": { "jsonPath": "$.queries[*]" }
      },
      "field": [
        {
          "@type": "cr:Field",
          "@id": "field/query_id",
          "name": "query_id",
          "description": "Stable query id (e.g. SS-0001).",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.query_id" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/setting",
          "name": "setting",
          "description": "Benchmark slice: single_skill, multi_skill_composition, distractor, outdated_redundant, budget_constrained.",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.setting" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/query",
          "name": "query",
          "description": "Natural-language user request (English and/or Korean).",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.query" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/difficulty",
          "name": "difficulty",
          "description": "Generator-assigned difficulty label.",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.difficulty" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/source",
          "name": "source",
          "description": "Query provenance (e.g. trigger_paraphrase).",
          "dataType": "sc:Text",
          "source": {
            "extract": { "jsonPath": "$.source" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/budget_tokens",
          "name": "budget_tokens",
          "description": "Token budget hint when setting is budget_constrained; null otherwise.",
          "dataType": "sc:Integer",
          "source": {
            "extract": { "jsonPath": "$.budget_tokens" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/gold_skills",
          "name": "gold_skills",
          "description": "Ground-truth skill_id list for the query.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "extract": { "jsonPath": "$.gold_skills[*]" }
          }
        },
        {
          "@type": "cr:Field",
          "@id": "field/distractor_skills",
          "name": "distractor_skills",
          "description": "Non-gold skill ids included as distractors when applicable.",
          "dataType": "sc:Text",
          "repeated": true,
          "source": {
            "extract": { "jsonPath": "$.distractor_skills[*]" }
          }
        }
      ]
    }
  ]
}