not analyzable — repository too large to scan in full
A memory of the codebase your AI agent can't get from grep: a persistent call graph, semantic search, and per-symbol review learnings over any repo. Self-hosted MCP server — Go, tree-sitter, Apache AGE.
Drift inferred · capture-to-capture
No drift recorded — single capability capture; advisories appear once its surface changes.
tools
-
call_trace
Trace call chains: callees (forward) or callers (reverse), with depth control
-
code_compare
Compare two repositories structurally: architecture, API design, code quality
-
code_graph
Query the persistent code knowledge graph in Apache AGE via natural language
-
code_health
Repo grade A-F: complexity, test coverage, dependency freshness, OSV vulnerabilities
-
code_research
BM25F + embeddings + DAG expansion for 10k+ file monorepos
-
code_search
Grep-like search with regex and path-glob filtering, context lines included
-
dataflow
IL/CFG taint tracking, dead stores, SQL/command-injection sinks
-
dead_code
Confidence-scored unused-symbol detection, not a flat list
-
debug_investigate
7-phase prod incident root cause: Prom spikes + Jaeger failed traces + symbol resolution + callgraph walks + LLM fusion + runtime binary drift, ranked to file:function
-
dep_graph
Build a dependency graph. Output as Mermaid, Graphviz DOT, or JSON
-
design_search
Find DESIGN.md systems by UI description (multilingual-e5-large, 1024-dim)
-
explore
Fast repo overview, no LLM, no clone for remote repos
-
file_parse
Parse a single file with tree-sitter. Returns symbol table or raw AST
-
find_duplicates
Find near-duplicate symbols in one repo via semantic similarity. Requires EMBED_URL + DATABASE_URL
-
fleet_versions
Diff pinned container-image references in source against deployed runtime containers. Catches the "config aligned, source looks right, behavior wrong" bug class
-
github_code_search
Search code on GitHub via the Code Search API. Returns file paths and matching fragments
-
impact_analysis
Configurable blast-radius depth (default 5). Direct callers, transitive callers, hotspot reordering by churn
-
prepare_change
Pre-change risk: impact analysis + dead-code check combined
-
remember_graph_insights
Persist a learning so it surfaces in future understand calls
-
repo_analyze
Analyze a repository. Deep mode (AST + LLM), quick mode (GitHub Code Search), or issue/PR search
-
repo_search
Discover repos across forges via parallel web search + GitHub/GitLab API, LLM-ranked
-
resolve_frame
Unminify a JS stack frame via source maps
-
review_delta
Differential blast radius between two git refs
-
review_pr
Differential blast radius between git refs; persists per-symbol learnings
-
rewrite
Structural AST search-replace with $WILDCARDS across 16 languages, dry-run and apply
-
semantic_search
Hybrid RRF: BM25F + pgvector + 1-hop AGE graph expansion. Find by concept, not keyword
-
site_analyze
Tech-stack and SEO audit, BFS crawler
-
site_crawl
BFS web crawler
-
suggest_reviewers
Rank reviewer candidates for a PR's file paths from authorship, co-change coupling, and recency
-
symbol_search
Search symbols (functions, types, consts) by name pattern across a repo
-
understand
Type-aware symbol deep-dive. Aggregates call_trace + symbol_search + complexity + tested_by + dead_code_score + prior learnings
-
wp_plugin_search
Search the WordPress.org plugin directory
last analysis: too-large
No code evidence — the analyzed source reached for no tracked permissions, tools, or hooks.