semantic-search — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited semantic-search (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
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.
L1 skill: pure function of L0 given a fixed embedding model. Same query + same corpus + same model → identical results, always. No LLM judgment. The returned message IDs are themselves the citations — no synthesis is performed.
This is the retrieval layer that feeds all L2 synthesis skills. When tunnel_state or what_do_i_think is called with a topic, it calls semantic-search first.
search-conversations) returns zero results but the topic probably was discussedquery: str # natural language query
limit: int = 10 # number of results to returnMarkdown list of the top N results, each carrying structural citations:
### [N] {conversation_title}
**Similarity**: 0.XXXX
**Source**: {source} · **Date**: YYYY-MM-DD · **conv_id**: {id}
> {400-char preview}Each result IS a citation. No synthesis. No interpretation. No merging.
what_do_i_think)dormant_contexts)tunnel_state)Internally wraps brain_mcp.server.tools_search.semantic_search(query, limit). Embedding generated via get_embedding(f"search_query: {query}"), searched against LanceDB message table, ranked by 1/(1+distance).
~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.