Stats
Actions
Tags
From beacon
Searches codebase using hybrid semantic embeddings + BM25 keyword + identifier boosting. Returns top matches with file paths, line ranges, scores, and code previews.
How this command is triggered — by the user, by Claude, or both
Slash command
/beacon:search-code <query> [query2] [--top-k N] [--threshold F]This command is limited to the following tools:
The summary Claude sees in its command listing — used to decide when to auto-load this command
# /search-code
Search the codebase using Beacon hybrid search (semantic embeddings + BM25 keyword matching + identifier boosting).
## Single query
1. Run `node ${CLAUDE_PLUGIN_ROOT}/scripts/search.js "$ARGUMENTS"`
2. Parse the JSON results — each match has `file`, `lines`, `similarity`, `score`, and `preview`
3. Review the `preview` field first — it often contains enough context to answer the query
4. Only read the actual source files if the preview is insufficient to answer the question
5. Summarize findings and cite file:line references
## Multi-query batch
Pass multiple quoted queries...Search the codebase using Beacon hybrid search (semantic embeddings + BM25 keyword matching + identifier boosting).
node ${CLAUDE_PLUGIN_ROOT}/scripts/search.js "$ARGUMENTS"file, lines, similarity, score, and previewpreview field first — it often contains enough context to answer the queryPass multiple quoted queries for a single HTTP round-trip:
node ${CLAUDE_PLUGIN_ROOT}/scripts/search.js "auth flow" "session handling"
Returns [{query, matches}, ...] grouped by query.
--top-k N — number of results (default: 10)--threshold F — minimum score cutoff (default: 0.35)--path <prefix> — scope results to files under a directory (e.g. --path src/middleware/)--no-hybrid — disable hybrid scoring, use pure vector searchfile — file pathlines — matched line range (e.g. "45-78")similarity — vector cosine similarityscore — final hybrid score (semantic + BM25 + identifier boost)preview — first 300 chars of matched code chunknpx claudepluginhub sagarmk/beacon-plugin/exploreExplores codebase iteratively by keyword: searches file names, code content, git history; refines results interactively or automatically up to depth N, tracks dependencies, summarizes structure and relations.
/exploreSearches the current codebase using a Gemini-powered Explore agent to find implementations, patterns, and code flows via natural language questions.
/researchResearches codebase using ChunkHound semantic analysis on a query. Outputs structured findings: overview, key components with file citations, relationships, and recommendations.
/coderlmExplores codebase structure using tree-sitter index: search symbols, trace callers/implementations, find tests, view file trees and scoped code via CLI tools.