[AdaWorldAPI]

Adds .claude/tools/codegraph_index.py and the generated .codegraph/ directory: compact, line-oriented, source-backed facts that any agent can search with rg. It is a projection: the code stays the truth. There is no new store, registry or executor.

Scope: the reduced slice EXO-230 gate 3 allows only with explicit approval.

• It covers Cargo dependencies, syntactic declarations and use leaves (cfg kept), and curated architecture rows.
• No Charon artifact exists for any real crate, so every crate reports CALL_GRAPH=ARTIFACT_MISSING, never an empty graph. Charon ingestion is tested against the Codegraph fixture.
• The original ~15 navigation commands were not recovered. The recipes in index.md are labelled NEW / DERIVED 2026-11-11.

What it adds

• codegraph_index.py: stdlib only, deterministic, with --write, --check and --self-test in the entries_index.py / supersession_index.py convention.
• .codegraph/: 108 generated files, about 4.9 MB:
  • TSVs: dependencies and reverse dependencies, symbols, imports, calls, decisions, gaps, freshness, coverage;
  • 90 crates/<crate>.index.md files;
  • 7 symbol views (ClassView, WideFieldMask, LaneBinding, MethodBinding, BindStore, LaneSource, FrameStore);
  • index.md, with the evidence legend and docs(.mermaid): architecture map of the DR / DT64 / mask-RISC stack (D-FOUNDRY-MAP-1) #1997's classDefs.
• .codegraph/architecture.tsv: the only hand-maintained input, 24 reviewed rows, each with path:line and an anchor token that --check re-validates. Architectural relationships come only from here, never from imports.
• Freshness: each crate has a digest over the git blob ids of its manifest and src/ files. --check names every stale crate. No timestamps, absolute paths or HEAD sha appear in the output.
• CI: .github/workflows/codegraph-index.yml runs the self-test and --check on PRs that touch the generator or .codegraph/. Whether to gate every crates/ change is left open on EXO-230.
• .claude/probes/2026-11-11-codegraph-mem-1/REPORT.md: crosswalk, negative tests, measurement, gap report and the suggested CLAUDE.md line (not applied).

Evidence

• Self-test: 11 falsifiers, all passing:
  • stale index;
  • cfg import and Cargo target cfg;
  • dispatch never CONFIRMED;
  • a plan never becomes shipped;
  • same-name symbols kept apart;
  • byte-identical output;
  • ARTIFACT_MISSING reported, not empty;
  • Charon fixture ingestion, including fnv64 STALE;
  • use-tree expansion;
  • anchor rot;
  • the lexer.
• Disable runs: 1, 2, 3, 4, 5, 7, 8 and 10 each turned red, then green again after restoring.
• On the real repo: appending a line to dt64/src/lane.rs makes --check report STALE crate lance-graph-dt64. Rebasing onto main after foundry-parity-lab: CanCook as a registered DT64 method (G6) + board record for #1991/#1995 #1996 made it flag two views before regeneration.
• A/B measurement: 10 fresh agents, 5 questions, graded against an answer key built from source beforehand. With the index first:
  • Architecture questions (Q1, Q4, Q5): correct, with 4.6k–8.3k fewer tokens and up to 4 fewer tool calls.
  • Census (Q2): no difference.
  • Call impact (Q3): worse, because calls are ARTIFACT_MISSING.
  • Breadth: source-only answers were broader.
  • Single runs per cell: an indication, not a result. Details are in the report.

What changes in existing behaviour

None. No Rust code changes; this adds a tool, generated files, a workflow and docs.

Open

• Real-crate Charon artifacts: EXO-229 gate A. Without them the index cannot answer call-impact questions.
• --check gating scope: whether it should gate every crates/ change (EXO-230).
• Import resolution: re-exports and globs are marked POSSIBLE, and tests, examples and benches are not scanned.

Board: EXO-230 (navigation), EXO-229 (intake).

---

Generated by Claude Code

[coderabbitai]

Important

• Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

Run configuration

• Source mode: Auto
• Repository: Unavailable
• Branch: Unavailable
• File: No configuration file from this repository applied
• Commit: Unavailable
• Layers: Organization UI
• Review profile: CHILL
• Plan: Advanced
• Run ID: 1d6642d7-b490-4a28-903a-3730e8b84b46

• Autofix · Keep fixing CodeRabbit findings and required CI, and resolving merge conflicts

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

Share

• X
• Mastodon
• Reddit
• LinkedIn

[chatgpt-codex-connector]

Codex Review Summary: ✅ Completed

This comment shows the latest Codex review activity on this pull request.

Review	Status	Commit	Review trigger
📝 Code Review	✅ Completed 2026-11-11T01:48:53.147589Z	faf7ce1	Draft marked ready

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review" or "@codex security review".

Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: faf7ce1021

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".