onsager-ai/dev-skills

codegraph

Pre-indexed code knowledge graph (MCP, SQLite + tree-sitter) for faster, lower-token exploration of brownfield codebases.

View source
Original skill document

Rendered from the source repository. Headings, examples, code, tables, links, and referenced images are preserved.

codegraph

Local code graph exposed as 8 MCP tools. Accelerates exploration with O(1) SQLite lookups and FTS5 search — slashing the grep + Read loops needed to orient in a codebase.

It does not replace grep. Codegraph is a tree-sitter approximation of the call graph, not a compiler-grade one: it is fast and low-token but unsound (it has false negatives). Treat it as the accelerator for orientation and traversal, and keep grep as the source of truth for completeness. See Soundness & limits below — this distinction is load-bearing, not a footnote.

When to use

  • Brownfield repo where structure is non-obvious
  • Task involves: "where is X used", "what breaks if I change Y", "trace flow from A to B"
  • Long-running session — cold-start cost amortizes over multiple queries

When NOT to use

  • Single-file edits or trivial lookups
  • Session likely shorter than cold-start cost (see references/spike.md for per-repo numbers)
  • Repo not initialized AND no cached .codegraph/ AND short session

Soundness & limits — read before trusting a result

Codegraph is built on tree-sitter, not the compiler, so its call/reference edges are an approximation. It fails in three confirmed ways. The governing rule falls out of them:

Grep is the source of truth for completeness. Codegraph is the accelerator for orientation and traversal. Never invert these.

The asymmetry is the whole argument: for a refactor, a false negative is dangerous (you miss a caller, ship a break); a false positive is merely annoying (you glance and discard). Codegraph has false negatives; grep does not. So codegraph may propose and rank, but grep confirms whenever the answer must be exhaustive.

The three failure modes (measured on a ~580-file Rust+TS repo):

  1. Unsound edges → false negatives. callers/callees/impact miss real call sites when the receiver type can't be resolved from syntax alone (calls on locals, generics, trait objects). Measured: a method with 8 real callers returned 2 — the 5 dropped were local.method() and generic-dispatch sites, with no warning. Exact on unique free-functions + direct calls (verified 1/1, 2/2); unreliable on method dispatch.
  2. Name resolution is fuzzy and overload-blind. Bare-name queries prefix-match (useWorkspace answered for useWorkspacesQuery — a different symbol) and conflate overloads (execute, with 19 definitions, returned sqlx's .execute() DB calls, not the trait method). For any name with multiple definitions, drop to codegraph_node on a specific symbol id; bare callers/callees are unattributable.
  3. The index lags the disk. CLI sees nothing until codegraph sync (~0.3s); the MCP file-watcher closes the gap to ~1–2s. Right after an edit, the index is wrong — grep/Read is ground truth until it catches up.

Decision matrix

TaskToolWhy
Orient / "explain subsystem X" / first surveycodegraph `explore`One call, ranked, verbatim source — far less context than grep+Read
Callers of a uniquely-named free functioncodegraph `callers` (trust it)Exact on unique names + direct calls
Find ALL callers before a signature change / renamegrep authoritative; codegraph only to pre-rankCodegraph is unsound here — completeness can't depend on it
Anything with an overloaded name (execute, lookup, run, new)grep, or codegraph node on a specific idBare-name codegraph conflates / under-resolves unpredictably
Impact of a core typecodegraph `impact --depth 1` to pre-rank, then grep to confirmDefault depth-2 inflates (test fns + one file entry per file) and isn't depth-ranked
Just-edited codegrep (or codegraph sync first)Index is stale until reindex
Trait-dispatch / generic call sitesgrep + readingBoth weak; codegraph silently drops them, grep at least surfaces the text

Workflow: start with explore for the map and callers/impact for direction — but the moment the answer must be complete (a rename, "did I get everyone?"), confirm with grep and treat any codegraph-vs-grep delta as "codegraph missed it," not "grep over-matched."

Prerequisites

  • Node 18+ on PATH
  • Project has .codegraph/ (run codegraph init if not)
  • For cloud sessions: cold-start cost verified via references/spike.md for the target repo

Files

  • references/install.md — installation for local / Claude Code Cloud / devcontainer
  • references/usage.md — which of the 8 MCP tools to use when, with main-session vs subagent rules
  • references/spike.md — cold-start measurement protocol; run once per new target repo before relying on codegraph in cloud sessions

Known integration targets

Cold-start measured locally (Node 22, 4 cores). Cloud cold-start still pending — re-run references/spike.md in a cloud session before relying on these numbers there.

RepoIndexed filesNodes / edgesDB sizeCold init+indexQuery wall
onsager-ai/onsager578 (rust + tsx/ts)8.6k / 20k24 MB12s~0.9s
crawlab-team/crawlab644 (go)6.8k / 15k11 MB6s~0.9s
codervisor/leanspec695 (tsx/rust/ts)8.9k / 21k23 MB10s~0.9s

Per-query wall is dominated by Node startup — the SQLite work itself is milliseconds. End-to-end cold-start including one-shot npx install is ~13–19s for repos in this size class.