vector-db-search — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited vector-db-search (Agent Skill) and scored it 87/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 3 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 3 flagged
The text {match} tells the agent to skip the normal "ask the user first" gate. Used adversarially it removes the human-in-the-loop check before destructive or sensitive actions, turning a normally-gated agent into a fire-and-forget executor.
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
This skill requires the chromadb and langchain packages defined in the plugin root.
Semantic (meaning-based) search against the ChromaDB vector store using a high-precision Parent-Child architecture. Use for Phase 2 of the 3-phase search protocol (RLM -> Vector -> Grep).
| Script | Role |
|---|---|
scripts/query.py | Semantic search CLI -- recovers context-rich parent chunks. |
scripts/operations.py | Core domain logic for retrieval. |
scripts/vector_config.py | Unified profile-based configuration loader. |
This skill defaults to In-Process mode for zero-latency direct disk access. No background server is required. This ensures maximum stability in isolated project environments.
Verify available profiles in .agent/learning/vector_profiles.json. The default profile is usually wiki.
Note: The --profile flag is mandatory to ensure the correct model and collection are loaded.
python ./scripts/query.py "your natural language question" --profile wiki --limit 5Results include ranked parent chunks (2,000 chars) that provide broad context to the LLM for reasoning.
--profile to ensure the correct semantic space is searched.query.py.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.