text
stringlengths
185
73.3k
repo
stringlengths
7
100
path
stringlengths
4
146
language
stringclasses
7 values
hash
stringlengths
16
16
score
float64
7
8.5
stars
int64
0
237k
"""One-off analyzer: compare OAuth Basic header + extract JS bundle across the Hymer and Eriba EHG XAPKs. Working tool, source/ is gitignored.""" from __future__ import annotations import base64 import re import sys import zipfile from pathlib import Path BASE = Path(__file__).resolve().parent ANDROID = BASE.parent /...
BetaHydri/hymer-connect-ha-ble
tools/analyze_ehg_apk.py
.py
06b9cfc77709f77f
7.48
8
"""Extract EHG component/device identifiers from the Hermes bundles and diff across app versions/brands to spot newly supported devices or mappings. Reads the bundles dumped by analyze_ehg_apk.py under source/androidapp/analysis/<app>/index.android.bundle plus the old app bundle in _archive_old_app/. source/ is gitign...
BetaHydri/hymer-connect-ha-ble
tools/analyze_ehg_devices.py
.py
603bbafd5265decb
7.48
8
"""Generate gated base.json sensor candidates from tools/ehg_metadata.json. Phase B (brandless auto-mapping): every EHG appliance is bound to a fixed bus, so a component mapped once in the universal base.json appears on any brand that reports it. This tool proposes `sensors` entries for the read-only slots of document...
BetaHydri/hymer-connect-ha-ble
tools/generate_base_from_metadata.py
.py
dcfe42b66ad5a044
7.48
8
"""Parse the committed EHG metadata Markdown into structured JSON. `docs/ehg-app-metadata.md` is the human-readable reference extracted from the decompiled EHG app. It holds: * a component table (bus -> component_id, kind, name, slot count, mapped?) * per-component slot tables under "### Bus <n> - ..." headers ...
BetaHydri/hymer-connect-ha-ble
tools/parse_ehg_metadata.py
.py
ddc300a031b9993d
7.48
8
#!/usr/bin/env python3 """Check every connector definition's field names against the live proxy. The connector proxy builds each command as ``command(**params)``, so a field name that the connector does not declare is silently never sent -- the profile value just does not arrive, and the failure surfaces as a confusin...
AOT-Technologies/m8flow
bin/check-connector-fields.py
.py
9de77c94d6893e03
7.42
6
#!/usr/bin/env python """Replace m8flow's copied model files with thin re-export shims. Background ---------- m8flow used to copy each upstream model file and add its tenant column, which placed LGPL-2.1 upstream code inside the Apache-2.0 tree. The schema delta now lives in m8flow_backend/models/tenant_schema.py, wh...
AOT-Technologies/m8flow
bin/generate-model-shims.py
.py
1a05dd140dd450b9
7.42
6
"""Connectors-tab page helpers (functional "page object"). Mirrors the style of :mod:`helpers.templates`: small functions that drive the Connectors view and assert on its stable ``data-testid`` locators. """ from __future__ import annotations import re from typing import Any from playwright.sync_api import Page, exp...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/connectors.py
.py
f56c0e384e6e01b6
7.92
6
import logging from playwright.sync_api import Page, expect, TimeoutError as PlaywrightTimeout from helpers.config import ( BASE_URL, API_PREFIX, DEFAULT_USERNAME, DEFAULT_PASSWORD, DEFAULT_TENANT, SUPER_ADMIN_USERNAME, SUPER_ADMIN_PASSWORD, MASTER_REALM_IDENTIFIER, KC_TIMEOUT, ...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/login.py
.py
1662275b5be52fb6
7.92
6
"""Shared helpers for navigating process groups and creating the test group.""" from __future__ import annotations import logging import re from playwright.sync_api import Page, TimeoutError as PlaywrightTimeout from helpers.config import BASE_URL, ELEMENT_TIMEOUT, PAGE_DATA_TIMEOUT, SHORT_TIMEOUT from helpers.wait...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/process_group_setup.py
.py
d6c074e9a0aa9fe6
7.92
6
import pytest from playwright.sync_api import Page, expect, TimeoutError as PlaywrightTimeout from helpers.config import BASE_URL, PAGE_DATA_TIMEOUT, ELEMENT_TIMEOUT, SHORT_TIMEOUT from helpers.waiters import wait_for_app_ready def navigate_to_templates(page: Page) -> None: """Click the Templates nav item and wai...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/templates.py
.py
ea2e9b55816eccbf
7.92
6
from playwright.sync_api import Page, expect from helpers.config import BASE_URL, PAGE_DATA_TIMEOUT, ELEMENT_TIMEOUT, SHORT_TIMEOUT from helpers.waiters import wait_for_app_ready def navigate_to_tenants(page: Page) -> None: """Navigate to the tenant management page.""" page.goto(f"{BASE_URL}/tenants") wai...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/tenants.py
.py
b85966c78812da81
7.92
6
"""Shared paths under ``test-results`` for screenshots and reports.""" from __future__ import annotations import hashlib import os from pathlib import Path try: from slugify import slugify as _slugify except ImportError: # pragma: no cover def _slugify(value: str) -> str: # type: ignore[misc] retu...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/test_artifacts.py
.py
ba3a336e379a38c4
7.92
6
from playwright.sync_api import Page, expect from helpers.config import APP_READY_TIMEOUT, PAGE_DATA_TIMEOUT def wait_for_app_ready(page: Page, timeout: int = APP_READY_TIMEOUT) -> None: """Wait until the m8flow app shell has fully loaded. Checks that both the user-actions menu and the SideNav logo are r...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/helpers/waiters.py
.py
cde25e1cdcfd1c68
7.92
6
"""Home inbox — tenant-admin vs super-admin header tabs and tenant column. CHK-01 and CHK-04: tenant-admin session. Super-admin counterparts live in ``roles/test_super_admin_home_tasks.py``. """ from __future__ import annotations import logging from playwright.sync_api import expect from helpers.config import ELEME...
AOT-Technologies/m8flow
extensions/m8flow-frontend/test/browser/home/test_home_rbac.py
.py
75cd6916afb49d76
7.92
6
#!/usr/bin/env python3 """Compare two LoCoMo benchmark runs side-by-side. Usage: python3 benchmarks/compare_runs.py run_a.json run_b.json python3 benchmarks/compare_runs.py run_a.json.conv0.jsonl run_b.json.conv0.jsonl Outputs a table comparing overall and per-category metrics, highlighting improvements and r...
star-ga/mind-mem
benchmarks/compare_runs.py
.py
bef6f9debe2f4549
7.6
15
#!/usr/bin/env python3 """Cross-Encoder A/B Test — retrieval-level comparison. Compares retrieval quality with and without cross-encoder reranking on LoCoMo conv-0. Measures: - Reciprocal Rank (MRR) of gold-answer keywords in retrieved context - Rank displacement per question (how many positions the best hit move...
star-ga/mind-mem
benchmarks/crossencoder_ab.py
.py
c4737284c370ada5
7.6
15
#!/usr/bin/env python3 """BM25F Field Weight Grid Search for mind-mem Recall Engine. Tests different BM25F field weight combinations against the LoCoMo retrieval benchmark, recording R@1, R@5, R@10, and MRR for each combination. Outputs a sorted comparison table and saves results to benchmarks/grid_search_results.json...
star-ga/mind-mem
benchmarks/grid_search.py
.py
15705596f4cb7ff8
7.6
15
#!/usr/bin/env python3 """LongMemEval Benchmark Harness for mind-mem recall engine. Evaluates mind-mem BM25 recall against the LongMemEval benchmark (ICLR 2025). Downloads the dataset from HuggingFace, converts chat sessions into mind-mem block format, runs retrieval queries, and reports Recall@K and MRR metrics with ...
star-ga/mind-mem
benchmarks/longmemeval_harness.py
.py
c35a5be962bc0bc7
7.6
15
#!/usr/bin/env python3 """Reproducible NIAH benchmark harness — emits independently-verifiable evidence. This runs the SAME Needle-In-A-Haystack code the test suite uses (imported from `tests/test_niah.py`, not reimplemented) and writes a third-party-reproducible artifact set: results.jsonl — one line per case (s...
star-ga/mind-mem
benchmarks/repro_niah.py
.py
fa4a03c511ed8ded
7.6
15
"""Shared pytest fixtures for mind-mem test suite.""" import os import shutil import sqlite3 import stat import sys import pytest # On Windows, pytest's `tmp_path` / `tmp_path_factory` fixtures use # `shutil.rmtree` for teardown. When a test has opened a SQLite # connection inside the temp dir (every mind-mem recall...
star-ga/mind-mem
conftest.py
.py
b4b364c71cbe6c16
8.1
15
"""The `gh pr view` fetch shared by Gauntlet campaign deciders.""" from __future__ import annotations import json import subprocess from pathlib import Path def _detail(exc: Exception) -> str: """One failure, described so the message is never EMPTY and always names what went wrong. ONLY THE TOTAL `except E...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/gh.py
.py
2326aeeef44e7081
7.64
18
"""Git ref selection shared by campaign base-fetch operations.""" from __future__ import annotations import hashlib import subprocess from dataclasses import dataclass @dataclass(frozen=True) class BaseFetchRefs: """The fully qualified refspec and exact local ref for one fetched base.""" refspec: str l...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/git_refs.py
.py
627364fe8806cd36
7.64
18
"""Shared JSONL object reader for Gauntlet campaign stores.""" from __future__ import annotations import json from collections.abc import Iterator from typing import cast class JsonlError(ValueError): """A malformed JSONL record, with its source line number.""" def object_lines(text: str) -> "Iterator[tuple[i...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/jsonl.py
.py
0d48c0ab06d1b1fc
7.64
18
"""Load sibling Gauntlet scripts whose filenames are not importable module names.""" from __future__ import annotations import importlib.util import sys from pathlib import Path from types import ModuleType def load_module_from_path(module_name: str, path: Path, *, register: bool = False) -> ModuleType | None: ...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/modules.py
.py
f34097e2d9666fdd
7.64
18
"""Shared mechanics for Gauntlet's source-mutation test harnesses.""" from __future__ import annotations import ast import re import types from collections.abc import Callable, Collection from pathlib import Path MarkedStatements = dict[str, tuple[str, ast.stmt]] ErrorFactory = Callable[[str], Exception] _MARKER_R...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/mutation.py
.py
3c686935562fc10c
7.64
18
"""Shared dispatch for the two public reviewer entry-point scripts.""" from __future__ import annotations import argparse import sys from pathlib import Path from types import ModuleType from typing import Protocol, cast from .modules import load_module_from_path class _ReviewPassOwner(Protocol): def add_emit_...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/review_door.py
.py
2aa6d80bc95d7461
7.64
18
"""Shared rendering for human-readable Gauntlet state tables.""" from __future__ import annotations import unicodedata def _hex_escape(ch: str) -> str: """``\\xNN`` for a byte-sized code point, ``\\uNNNN`` above it.""" return f"\\x{ord(ch):02x}" if ord(ch) < 0x100 else f"\\u{ord(ch):04x}" def escape_cell(...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/_gauntlet/table.py
.py
0d7dab07f44b0331
7.64
18
#!/usr/bin/env python3 """Refuse to format a file whose write could land OUTSIDE the worktree. The cheap CI-fix subagent runs a formatter (`gofmt -w`, …) that writes bytes back through the path it is given. A formatter writes THROUGH a symlink: point one at a file elsewhere on the machine and the bytes land there, whi...
lestrrat-ai/claude-code-plugins
plugins/gauntlet/skills/campaign/scripts/format-preflight.py
.py
3aa9b0cf501b5065
7.64
18
"""LocalChat application entry point. Production (Docker / Uvicorn):: uvicorn "app:create_uvicorn_app" --factory --host 0.0.0.0 --port 5000 Development:: python app.py """ import os import socket import subprocess import sys import time from pathlib import Path # Add src to path if running from root sys....
jwvanderstam/LocalChat
app.py
.py
d59fb03d1daebe09
7.42
6
"""Expand the runtime environment variables a shell would have expanded. The hardened base image ships no shell, so the previous `sh -c` CMD — and the `${SERVER_PORT:-5000}` style defaults inside it — cannot run. This does that job in Python and then `exec`s uvicorn in place, so the server stays PID 1 and signals reac...
jwvanderstam/LocalChat
docker-entrypoint.py
.py
cd0353e2b2cddeb8
7.42
6
""" MCP Server Base =============== Lightweight JSON-RPC 2.0 server base class for LocalChat MCP domain servers. Each server exposes tools via POST /mcp and a GET /health endpoint. Protocol: - tools/list -> {"jsonrpc":"2.0","id":N,"method":"tools/list","params":{}} - tools/call -> {"jsonrpc":"2.0","id":N,"metho...
jwvanderstam/LocalChat
mcp_servers/base.py
.py
b847c8604b8a77a5
7.42
6
"""Alembic environment — connects to the app's PostgreSQL via src.config.""" from __future__ import annotations import urllib.parse from logging.config import fileConfig from alembic import context from sqlalchemy import create_engine alembic_cfg = context.config if alembic_cfg.config_file_name is not None: # ...
jwvanderstam/LocalChat
migrations/env.py
.py
9f28a81d99469163
7.42
6
"""Early additive columns — conversations, documents, conversation_messages. Adds columns that were introduced during initial feature development: conversations.document_ids documents.content_hash, doc_type, chunker_version, local_only conversation_messages.plan_json conversations.memory_extracted_at All stat...
jwvanderstam/LocalChat
migrations/versions/0002_early_additive_columns.py
.py
a665200ffb4a2ed0
7.42
6
"""v1.1/v1.5 documents columns — language, last_ingested_at, source_id. Adds: documents.language (v1.1 — multilingual detection) documents.last_ingested_at (v1.1 — scheduled re-ingest tracking) + backfill documents.source_id (v1.5 — multi-source connector ID) + index All statements are idempotent (IF NOT EXISTS...
jwvanderstam/LocalChat
migrations/versions/0004_documents_language_ingest_source.py
.py
ea3e620506edbf5c
7.42
6
"""CW-1: soft-delete columns for documents and document_chunks. Adds: documents.deleted_at (TIMESTAMPTZ) — set on soft-delete; NULL = live documents.deleted_by (UUID FK → users.id) — who triggered the retirement document_chunks.deleted_at (TIMESTAMPTZ) — reserved for per-chunk retirement All statements are ...
jwvanderstam/LocalChat
migrations/versions/0005_document_soft_delete.py
.py
866b563b4be86fae
7.42
6
"""CW-2a: soft-delete columns for conversations. Adds: conversations.deleted_at (TIMESTAMPTZ) — set on soft-delete; NULL = live conversations.deleted_by (UUID FK → users.id) — who triggered the retirement All statements are idempotent (IF NOT EXISTS). Revision ID: 0006 Revises: 0005 Create Date: 2026-06-28 """...
jwvanderstam/LocalChat
migrations/versions/0006_cw2a_conversations_soft_delete.py
.py
c8a7136b9ba83761
7.42
6
"""CW-2b: soft-delete columns for users. Adds: users.deleted_at (TIMESTAMPTZ) — set on soft-delete; NULL = live users.deleted_by (UUID FK → users.id, self-referential) — admin who retired the account All statements are idempotent (IF NOT EXISTS). Revision ID: 0007 Revises: 0006 Create Date: 2026-06-28 """ from...
jwvanderstam/LocalChat
migrations/versions/0007_cw2b_users_soft_delete.py
.py
dbdedf4f20b68d4c
7.42
6
"""Enforce one live document per (filename, workspace_id). document_exists() previously ignored workspace_id entirely, so two workspaces uploading a same-named file could read and soft-delete each other's document. Now that the read is workspace-scoped, this migration closes the other half: nothing previously stopped ...
jwvanderstam/LocalChat
migrations/versions/0013_documents_unique_filename_workspace.py
.py
c4945077e5f7bac8
7.42
6
"""RBAC-1 prerequisite: backfill workspace_members so nobody is locked out. Membership was never written on the creation path, so on an existing instance no user is a member of any workspace. Once RBAC-1 enforces membership, that state denies everyone everything. This backfill establishes the starting membership. Two...
jwvanderstam/LocalChat
migrations/versions/0014_rbac1_backfill_workspace_members.py
.py
97741a47ded07a4c
7.42
6
"""Workspace API keys — programmatic access to one workspace, no user attached.""" from alembic import op revision = "0015" down_revision = "0014" branch_labels = None depends_on = None def upgrade() -> None: op.execute(""" CREATE TABLE IF NOT EXISTS workspace_api_keys ( id UUID PRI...
jwvanderstam/LocalChat
migrations/versions/0015_workspace_api_keys.py
.py
04c65fb84b67e6a5
7.42
6
"""PERF-2 — measure /api/chat under concurrent SSE load. PERF-1's defect (sync retrieval inline on the event loop, so one slow query stalled every other request) survived for months because nothing ever measured behaviour with more than one client. A single-user stopwatch cannot see it: the request that blocks the loo...
jwvanderstam/LocalChat
scripts/bench_concurrency.py
.py
5f26c280f5318fab
7.42
6
"""DEL-2 — measure retrieval quality against a fixed set of question/source pairs. The ticket asks whether GraphRAG's 1-hop expansion earns its place. That is not a question inspection can answer, and it is not a question a single anecdote can answer either: it needs the same questions asked of the same corpus with th...
jwvanderstam/LocalChat
scripts/eval_retrieval.py
.py
afbb89914070c5cb
7.42
6
"""TQ-3 — the nightly mutation gate. Runs `mutmut` over the isolation-critical modules and fails when a module's kill rate falls under the agreed threshold. Coverage says a line executed; this says something asserted its behaviour. Scoped ruthlessly on purpose. A whole-repo run is hours, and the modules here are wher...
jwvanderstam/LocalChat
scripts/mutation_gate.py
.py
78347147205704e8
7.42
6
""" Model Registry ============== Maps logical model classes (FAST, BASE, LARGE, CODE, VISION) to concrete Ollama model IDs and request parameters. Model IDs are read from environment variables at import time. An empty model_id means "use the user-selected active model" — the registry never overrides unless a specif...
jwvanderstam/LocalChat
src/agent/models.py
.py
2c179d7165c8298e
7.42
6
""" Agent Result Types ================== Data classes for the output of AggregatorAgent.run(). Kept in a separate module so they can be imported without pulling in the full agent machinery (useful in tests and type annotations). """ from __future__ import annotations from dataclasses import dataclass, field @dat...
jwvanderstam/LocalChat
src/agent/result.py
.py
3e32229d3c8e8378
7.42
6
""" Base Connector Interface ======================== All live-sync connectors implement ``BaseConnector``. The interface is intentionally synchronous so connectors can run inside a normal thread pool without requiring an async event loop in the main application process. """ from __future__ import annotations impor...
jwvanderstam/LocalChat
src/connectors/base.py
.py
78ae1704b1073d67
7.42
6
#!/usr/bin/env python3 """Check OpenAPI spec for breaking changes using oasdiff. This script compares two OpenAPI specs and detects breaking changes, returning structured JSON output for consumption by CI or local tooling. Usage: ./check-breaking-changes.py --base devel --head HEAD ./check-breaking-changes.py...
syntara-orchestration/syntara
backend/scripts/openapi/check-breaking-changes.py
.py
a9f31106d8c0df84
7.48
8
#!/usr/bin/env python3 """Post or update GitHub PR comment for breaking changes check results. This script formats the breaking changes check results and posts/updates a PR comment using the GitHub API via gh CLI. Usage: ./post-breaking-changes-comment.py --results results.json --pr-number 123 ./post-breaking...
syntara-orchestration/syntara
backend/scripts/openapi/post-breaking-changes-comment.py
.py
e1bc1921b81965e5
7.48
8
#!/usr/bin/env python3 """Post or update GitHub PR comment for contract regeneration check results. This script formats the contract check results and posts/updates a PR comment using the GitHub API via gh CLI. Usage: ./post-contract-regeneration-comment.py --results results.json --pr-number 123 ./post-contra...
syntara-orchestration/syntara
backend/scripts/openapi/post-contract-regeneration-comment.py
.py
a7c9a1ad952b66f3
7.48
8
import pytest from unittest.mock import MagicMock from homeassistant.core import HomeAssistant from homeassistant.config_entries import ConfigEntry from custom_components.power_max_tracker.coordinator import PowerMaxCoordinator from custom_components.power_max_tracker.const import ( CONF_SOURCE_SENSOR, CONF_M...
perosb/power_max_tracker
conftest.py
.py
81aa51d19b6e1232
8.16
20
import uuid import voluptuous as vol from homeassistant import config_entries from homeassistant.helpers import selector from .const import ( DOMAIN, CONF_SOURCE_SENSOR, CONF_MONTHLY_RESET, CONF_NUM_MAX_VALUES, CONF_BINARY_SENSOR, CONF_PRICE_PER_KW, CONF_POWER_SCALING_FACTOR, CONF_START_...
perosb/power_max_tracker
custom_components/power_max_tracker/config_flow.py
.py
d9858b32f2effe03
7.66
20
"""Tests for PowerMaxTracker config flow. Note: These tests require a full Home Assistant development environment with all dependencies. They will fail when run in a standalone environment without HA installed. """ import pytest from unittest.mock import MagicMock, AsyncMock, patch from homeassistant.core import Hom...
perosb/power_max_tracker
custom_components/power_max_tracker/tests/test_config_flow.py
.py
b89ac3d4bad5849c
7.16
20
"""Tests for PowerMaxCoordinator helper methods.""" import pytest from datetime import datetime from unittest.mock import MagicMock # Test the helper method logic without importing the full coordinator # We'll simulate the helper methods here for testing def watts_to_kilowatts(watts: float) -> float: """Convert...
perosb/power_max_tracker
custom_components/power_max_tracker/tests/test_coordinator_helpers.py
.py
589b66624cf62113
8.16
20
"""Tests for PowerMaxTracker __init__.py services. Note: These tests require a full Home Assistant development environment with all dependencies. They will fail when run in a standalone environment without HA installed. """ import pytest from unittest.mock import MagicMock, AsyncMock, patch from homeassistant.const ...
perosb/power_max_tracker
custom_components/power_max_tracker/tests/test_init.py
.py
666ee7e6684057aa
8.16
20
#!/usr/bin/env python3 """Generate Claude-native subagent type definitions from the models.conf catalog. `CFG_NATIVE_AGENT_CATALOG` (adapters/claude/config/models.conf) maps each native subagent type to one portable execution profile. This generator resolves every profile through `utilities/model_profile.py` and write...
dmlguq456/hearting
adapters/claude/bin/sync-native-agents.py
.py
f7d8daf7e7e52cf3
7.45
7
#!/usr/bin/env python3 """Generate the Claude-native plugin projection for the portable harness. Claude is the native runtime — skills/agents live at the repo-root SoT directly (`adapters/claude/skills/`, `adapters/claude/agents/`), unlike Codex which needs a sync-native-skills generator first. This is the *first* Cla...
dmlguq456/hearting
adapters/claude/bin/sync-native-plugin.py
.py
58147b4b43d922ba
7.45
7
import logging from collections import deque from datetime import datetime from typing import Dict, Any, List, Optional from loguru import logger # Ring buffer for retaining recent Cloud Run / backend logs for UI diagnostics DIAGNOSTIC_LOG_BUFFER: deque = deque(maxlen=1500) # Structured storage for individual turn la...
manishkjs/gemini_live_pipecat
server/diagnostic_buffer.py
.py
8f1e5d4a50ce8b51
7.57
13
from loguru import logger from pipecat.processors.frame_processor import FrameProcessor, FrameDirection from pipecat.frames.frames import ( Frame, TextFrame, InterruptionFrame, LLMMessagesAppendFrame, LLMFullResponseEndFrame, TranscriptionFrame, ) class RepeatOnInterruptionProcessor(FrameProce...
manishkjs/gemini_live_pipecat
server/processors/repeat_on_interruption.py
.py
e52db082483f26f7
7.57
13
import asyncio import os import re from loguru import logger from pipecat.services.llm_service import FunctionCallParams from pipecat.adapters.schemas.function_schema import FunctionSchema try: import vertexai from vertexai.preview import rag except ImportError: vertexai = None rag = None import google....
manishkjs/gemini_live_pipecat
server/rag_function.py
.py
828c5819e38cc950
7.57
13
import unittest import os import sys server_dir = os.path.dirname(os.path.dirname(os.path.abspath(__file__))) if server_dir not in sys.path: sys.path.insert(0, server_dir) from agent import validate_stt_model, validate_llm_model, validate_tts_model class TestModelRouting(unittest.TestCase): def test_stt_mod...
manishkjs/gemini_live_pipecat
server/tests/test_model_routing.py
.py
c0e0633aa56033df
7.07
13
"""tasks:// adapter — reference example of a third-party reveal plugin. Demonstrates the plugin discovery mechanism (BACK-256): dropped into <project>/.reveal/adapters/tasks/, this package is auto-discovered on `cd` into the project, no installation or reveal-core changes required. See this example's README.md for a w...
Semantic-Infrastructure-Lab/reveal
examples/plugin-adapters/tasks-adapter/.reveal/adapters/tasks/adapter.py
.py
f78361af30de566b
7.52
10
"""File analysis and structure extraction for AST adapter.""" import builtins as _builtins_module import os import sys from pathlib import Path from typing import Dict, List, Any, Optional from ...utils.path_utils import is_skippable_dir from .call_graph import build_symbol_map, resolve_callees # All public names i...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/analysis.py
.py
81c6965a6b9295ae
7.52
10
"""Cross-file call graph resolution for the AST adapter (Phase 3). Given a function's raw `calls` list (bare names like ["validate_item", "db.insert"]), and the import map for the file, resolves entries to their source files where possible. Keeps `calls` as List[str] (backward-compat) and adds `resolved_calls` as Lis...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/call_graph.py
.py
dbfb28ad329128cc
7.52
10
"""Filter matching logic for AST adapter.""" import re from fnmatch import fnmatch from typing import Dict, List, Any, Set from ...utils.query import compare_values # Filter keys `matches_filters` special-cases (mapped to a differently-named # or derived element field, not a literal dict key) — always valid filter # ...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/filtering.py
.py
56e0b4044a692163
7.52
10
from typing import Any, Callable, Dict, List, Optional from .nav_exits import collect_deps from .nav_effects import ( collect_effects, collect_effects_transitive, format_effect_target, render_effects_transitive, ) _PHP_SUPERGLOBALS = frozenset({ '$_GET', '$_POST', '$_SESSION', '$_SERVER', '$_FILE...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_boundary.py
.py
1d82d8a0250cb201
7.52
10
"""Tree-sitter contract extraction for C++ — abstract classes and their subclasses. BACK-403 pt 2 (contracts breadth). C++ has no `interface` keyword; the idiomatic contract is an **abstract class** — a `class`/`struct` with at least one *pure virtual* method (`virtual T f() = 0;`). Implementors are declared **explici...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_contracts_cpp.py
.py
6f4b94d1a14e01ad
7.52
10
"""Tree-sitter contract extraction for Go — interfaces and their implementers. BACK-403 pt 2 (contracts breadth). Go's contract construct is the **interface** (`type Foo interface { ... }`), the public-API / moat surface a DD read wants. Two things make Go need its own scanner rather than the shared `_scan_contracts_t...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_contracts_go.py
.py
1d0c8182efaab3eb
7.52
10
"""Tree-sitter contract extraction for Rust — traits and their implementors. BACK-403 pt 2 (contracts breadth). Rust's contract construct is the **trait** (`trait Foo { ... }`); a type satisfies it through an **explicit** `impl Foo for Bar` block — so, unlike Go's implicit method-set satisfaction, implementors are *de...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_contracts_rust.py
.py
6faecd11b83da143
7.52
10
"""Loop-focused navigation: collect_loops (--loopmap), collect_fanout (--fanout). BACK-439b: agents routinely need "which loops exist, what do they iterate, and which side effects happen inside them" for N+1 database checks, per-item HTTP calls, filesystem fan-out, and retry-safety review. The raw facts already exist ...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_loops.py
.py
0ecfa23f4744924f
7.52
10
"""Persistent/shared-state mutation surface: collect_statewrites (--statewrites). BACK-439c: --mutations is local-variable/refactor oriented (read-after-write hazards); --sideeffects is call-oriented. Neither answers "what shared state does this code mutate" — dogfood-confirmed blind spot on GitRefResource._parse_and_...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_statewrites.py
.py
88eba68fef4a0645
7.52
10
"""Tree-sitter surface extraction for C++ — env vars, FS writes, HTTP routes, CLI, includes. BACK-403 pt 2 (surface breadth). C++ has no single dominant web framework, so route coverage is the two most common shapes, both leading-`/` guarded: - **cpp-httplib**: `svr.Get("/path", handler)` — a `field_expression` call ...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_surface_cpp.py
.py
0b306f58d1baa76c
7.52
10
"""Tree-sitter surface extraction for C# — env vars, FS writes, HTTP routes, CLI, imports. BACK-403 pt 2. Mirrors nav_surface_java.py's shape: attributes for ASP.NET Core HTTP routes, `Environment.GetEnvironmentVariable` for env access, `static void/int Main` for the CLI entrypoint, and using-root taxonomy for network...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_surface_csharp.py
.py
3a8f4049c2d224dc
7.52
10
"""Tree-sitter surface extraction for Go — env vars, FS writes, HTTP routes, CLI, imports. BACK-403 pt 2 (surface breadth). Mirrors nav_surface_java.py's categorised-dict shape but walks Go's grammar: - **HTTP routes** for the dominant Go routers, all of which express a route as a method call whose selector field i...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_surface_go.py
.py
d28bbaeb77932004
7.52
10
"""Tree-sitter surface extraction for PHP — env vars, HTTP routes, imports. BACK-403 pt 2 (surface breadth). Mirrors nav_surface_java.py's categorised-dict shape but walks PHP's grammar for the three dominant web frameworks: - **Laravel** routes: ``Route::get('/path', ...)`` — a ``scoped_call_expression`` whose rec...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/nav_surface_php.py
.py
211cc56a02cf2601
7.52
10
"""Query parsing and formatting for AST adapter.""" from typing import Dict, Any from ...utils.query import parse_query_filters def parse_equality_value(key: str, value: str) -> Dict[str, Any]: """Parse equality parameter value based on content. Args: key: Parameter key (e.g., 'type', 'name') ...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/queries.py
.py
c4f5128455c2ad50
7.52
10
"""Rendering for AST query adapter.""" import sys class AstRenderer: """Renderer for AST query results.""" @staticmethod def render_structure(result: dict, format: str = 'text') -> None: """Render AST query results. Args: result: Query result dict from AstAdapter.get_structu...
Semantic-Infrastructure-Lab/reveal
reveal/adapters/ast/renderer.py
.py
d6c97d38227de58e
7.52
10
"""hermes-talk — OpenAI Realtime speech-to-speech voice for Hermes Agent. ``register(ctx)`` wires five surfaces: the ``hermes talk`` CLI command, the ``/talk`` slash command, lifecycle hooks (session end plus the v0.6 subagent start/stop pair that powers push-based run control), and (when the host exposes the provider...
TheSmokeDev/hermes-talk
__init__.py
.py
fba3a37d91884ece
7.66
20
"""OpenAI Platform auth resolution with Codex OAuth fallback. Port of the proven Talk Mode auth ordering (itself a port of OpenClaw PR #100671, "Reuse Codex OAuth for OpenAI Realtime voice"). Resolution order, fail-closed at every step: 1. ``TALK_OPENAI_API_KEY`` — a Talk-scoped configured key. Present but blank f...
TheSmokeDev/hermes-talk
talk_auth.py
.py
1a92b28311227042
7.66
20
"""Live capability catalog — what this Hermes session can ACTUALLY do. Talk could always describe itself; it could never check. Asked "what can you do right now?" the model either recited its system prompt or spent a whole delegated agent turn finding out. This module is the third answer: a bounded, read-only snapshot...
TheSmokeDev/hermes-talk
talk_capabilities.py
.py
fe1c867650e1bf7e
7.66
20
"""Transport-neutral admission primitives for canonical Talk sessions.""" from __future__ import annotations import asyncio import uuid from collections.abc import Callable from dataclasses import dataclass from typing import Any try: from . import talk_config, talk_core_realtime except ImportError: # pragma: n...
TheSmokeDev/hermes-talk
talk_core_session.py
.py
e76f486eb148407e
7.66
20
"""Realtime session instructions — the voice preamble and identity assembly. A Realtime session prompt is re-read on EVERY turn, so the per-section caps here are a budget, not a preference. The preamble is the behavioural contract and always ships; host identity sections are optional and additive. Fail-open is the ru...
TheSmokeDev/hermes-talk
talk_identity.py
.py
91d37a87dd17710b
7.66
20
"""Subagent lifecycle — push-based, from the host's own hook bus. v0.5 learned that a child was gone by PULLING: ``degrade_gone_children()`` sweeps the delegation registry at ``check_work`` time, so a finished child went unnoticed until the operator next asked. The 0.20 host fires ``subagent_start`` / ``subagent_stop`...
TheSmokeDev/hermes-talk
talk_lifecycle.py
.py
618dfe512cea1602
7.66
20
"""OpenAI TTS and transcription providers for Hermes's pluggable backends. Bonus surface, not the point of this plugin: hermes-talk already resolves an OpenAI Platform key, so it can also service Hermes's turn-based speech hooks. The ABCs are imported defensively — the plugin must import and test green on a machine wi...
TheSmokeDev/hermes-talk
talk_providers.py
.py
1a146de047d68dc3
7.66
20
"""Vault recall — the durable-notes lookup a voice session can actually make. Hermes's memory PROVIDERS publish a system-prompt block that tells the model which tools to call (``homie_memory_search``, ``homie_memory_context``, …). Those tools exist in a text agent's registry and **not** in a Realtime session's, so pas...
TheSmokeDev/hermes-talk
talk_vault.py
.py
7016e83fecee68d5
7.66
20
"""Pure OpenAI Realtime wire layer — session payloads and ephemeral mints. Ported from a prior proven Talk Mode implementation. This module knows the OpenAI Realtime wire format and NOTHING about the host: no Hermes imports, no identity assembly, no tool handlers. Callers pass instructions, tools, and an auth token in...
TheSmokeDev/hermes-talk
talk_wire.py
.py
1fd3df594eeb6d13
7.66
20
"""Shared suite plumbing.""" from __future__ import annotations import pytest @pytest.fixture(autouse=True) def _ephemeral_runs_optin(monkeypatch): """The suite's EXPLICIT opt-in to non-durable run acceptance. The run-history tee is inert under pytest by design (see ``talk_runs._history_enabled``), and...
TheSmokeDev/hermes-talk
tests/conftest.py
.py
880d6eb3b186a9ab
7.16
20
"""Load adversarial test payloads from ``tests/fixtures/``. The strings these helpers return are attack-shaped bytes — injection text, destructive commands, dummy credentials — quoted by the containment and redaction tests to prove those protections hold against the real thing. They live in ``.fixture`` files because ...
TheSmokeDev/hermes-talk
tests/fixture_data.py
.py
20588dd33a2084bc
8.16
20
"""Dashboard manifest — the host reads this, so a typo is a silent dead tab. ``_discover_dashboard_plugins`` swallows a bad manifest with a log warning and moves on, and ``serve_plugin_asset`` 404s a missing entry file. Neither failure reaches the operator as anything but an empty tab, so the shape is asserted here in...
TheSmokeDev/hermes-talk
tests/test_dashboard_manifest.py
.py
73eb557926c97ee8
8.16
20
"""Helpers for lazy public exports and optional dependency errors.""" from __future__ import annotations from collections.abc import Mapping from importlib import import_module from typing import Final, NoReturn LazyExports = Mapping[str, tuple[str, str]] # Top-level import names installed by each optional extra's ...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/_imports.py
.py
cf85b9a6623b5fb9
7.45
7
"""Clipboard support for interactive CLI flows. Uses a local tool when one can serve the session; otherwise OSC 52 so the terminal sets the clipboard even over SSH and tmux. """ from __future__ import annotations import base64 import os import platform import shutil import subprocess import sys _LOCAL_TOOL_TIMEOUT ...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/clipboard.py
.py
4816380a1821793b
7.45
7
"""Benchmark catalog and run management commands.""" from __future__ import annotations from pathlib import Path import typer from osmosis_ai.cli.options import ( all_option, cursor_option, limit_option, log_limit_option, ) from osmosis_ai.cli.output import CommandResult app: typer.Typer = typer.Ty...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/benchmark.py
.py
29ded4fa8fbc029d
7.45
7
"""Dataset management commands (thin shell delegating to platform/cli/dataset.py).""" from __future__ import annotations import typer from osmosis_ai.cli.options import ( all_option, cursor_option, limit_option, log_limit_option, ) from osmosis_ai.cli.output import CommandResult app: typer.Typer = t...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/dataset.py
.py
f3e9b1608dfaf8c1
7.45
7
from __future__ import annotations from typing import NoReturn import typer from osmosis_ai.cli.options import all_option, limit_option from osmosis_ai.cli.output import CommandResult from osmosis_ai.platform.constants import MAX_LOG_PAGE_SIZE app: typer.Typer = typer.Typer( help="Manage a remote rollout server...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/dev/server.py
.py
c1a74d3b6b73295b
7.45
7
"""Model management commands (thin shell delegating to platform/cli/model.py). Models cover both base (foundation) models and LoRA models produced by training runs: osmosis model list -> GET /api/cli/models/base + /api/cli/models/lora osmosis model info <lora-model> -> GET /api/...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/model.py
.py
8b794386dfa3d2c7
7.45
7
"""Rollout commands: list.""" from __future__ import annotations from typing import Annotated import typer from osmosis_ai.cli.options import all_option, limit_option from osmosis_ai.cli.output import CommandResult app: typer.Typer = typer.Typer( help="Manage rollouts (init, list).", no_args_is_help=True, ...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/rollout.py
.py
056baecc3474ef3c
7.45
7
"""Secret management commands. ``osmosis secret`` manages secrets — the same records referenced by the ``[secrets]`` section in submit configs. Evaluation configs must include this section; default OpenAI eval configs should include ``OPENAI_API_KEY`` and use ``required = []`` only when no secret refs are needed. Trai...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/secret.py
.py
dd64bebc52d555c4
7.45
7
"""``osmosis template`` command shell. Thin Typer wrapper that delegates to :mod:`osmosis_ai.templates.cli`. Heavy imports stay inside the command bodies per the CLI lazy-loading contract. """ from __future__ import annotations from typing import Any import typer from osmosis_ai.cli.output import CommandResult ap...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/template.py
.py
db24a83362cef1a7
7.45
7
"""Training run management commands (thin shells delegating to platform/cli/train.py).""" from __future__ import annotations from pathlib import Path import typer from osmosis_ai.cli.options import ( all_option, cursor_option, limit_option, log_limit_option, ) from osmosis_ai.cli.output import Comma...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/commands/train.py
.py
b27d226e6a900efc
7.45
7
"""Console output facade with Rich and output-context aware rendering. Rich automatically strips ANSI control codes when output is not directed to a terminal (e.g., piped to a file), and respects the NO_COLOR environment variable. Rich is imported on first use in rich mode so ``--json`` / ``--plain`` paths can import...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/console.py
.py
0f14d28dd7464b09
7.45
7
from __future__ import annotations from collections.abc import Mapping from enum import StrEnum from typing import Any class CLIErrorCode(StrEnum): """Public CLI error codes in JSON error envelopes and ``CLIError.code``.""" VALIDATION = "VALIDATION" AUTH_REQUIRED = "AUTH_REQUIRED" NOT_FOUND = "NOT_F...
Osmosis-AI/osmosis-sdk-python
osmosis_ai/cli/errors.py
.py
310f9b5c16431350
7.45
7