GitNexus — Code Intelligence

This project is indexed by GitNexus as Claude-cod-top-2026 (2586 symbols, 5616 relationships, 47 execution flows). Use the GitNexus MCP tools to understand code, assess impact, and navigate safely.

If any GitNexus tool warns the index is stale, run npx gitnexus analyze in terminal first.

Always Do

• MUST run impact analysis before editing any symbol. Before modifying a function, class, or method, run gitnexus_impact({target: "symbolName", direction: "upstream"}) and report the blast radius (direct callers, affected processes, risk level) to the user.
• MUST run gitnexus_detect_changes() before committing to verify your changes only affect expected symbols and execution flows.
• MUST warn the user if impact analysis returns HIGH or CRITICAL risk before proceeding with edits.
• When exploring unfamiliar code, use gitnexus_query({query: "concept"}) to find execution flows instead of grepping. It returns process-grouped results ranked by relevance.
• When you need full context on a specific symbol — callers, callees, which execution flows it participates in — use gitnexus_context({name: "symbolName"}).

When Debugging

1. gitnexus_query({query: "<error or symptom>"}) — find execution flows related to the issue
2. gitnexus_context({name: "<suspect function>"}) — see all callers, callees, and process participation
3. READ gitnexus://repo/Claude-cod-top-2026/process/{processName} — trace the full execution flow step by step
4. For regressions: gitnexus_detect_changes({scope: "compare", base_ref: "main"}) — see what your branch changed

When Refactoring

• Renaming: MUST use gitnexus_rename({symbol_name: "old", new_name: "new", dry_run: true}) first. Review the preview — graph edits are safe, text_search edits need manual review. Then run with dry_run: false.
• Extracting/Splitting: MUST run gitnexus_context({name: "target"}) to see all incoming/outgoing refs, then gitnexus_impact({target: "target", direction: "upstream"}) to find all external callers before moving code.
• After any refactor: run gitnexus_detect_changes({scope: "all"}) to verify only expected files changed.

Never Do

• NEVER edit a function, class, or method without first running gitnexus_impact on it.
• NEVER ignore HIGH or CRITICAL risk warnings from impact analysis.
• NEVER rename symbols with find-and-replace — use gitnexus_rename which understands the call graph.
• NEVER commit changes without running gitnexus_detect_changes() to check affected scope.

Tools Quick Reference

Impact Risk Levels

Resources

Self-Check Before Finishing

Before completing any code modification task, verify:

1. gitnexus_impact was run for all modified symbols
2. No HIGH/CRITICAL risk warnings were ignored
3. gitnexus_detect_changes() confirms changes match expected scope
4. All d=1 (WILL BREAK) dependents were updated

Keeping the Index Fresh

After committing code changes, the GitNexus index becomes stale. Re-run analyze to update it:

npx gitnexus analyze

If the index previously included embeddings, preserve them by adding --embeddings:

npx gitnexus analyze --embeddings

To check whether embeddings exist, inspect .gitnexus/meta.json — the stats.embeddings field shows the count (0 means no embeddings). Running analyze without --embeddings will delete any previously generated embeddings.

Claude Code users: A PostToolUse hook handles this automatically after git commit and git merge.

CLI

Project Agent Rules — Claude-cod-top-2026

This section covers project-specific rules for all AI agents (Claude Code subagents, Dispatch, Codex, etc.). GitNexus impact rules above take priority for all code edits.

Read First (every agent, in order)

1. .claude/memory/activeContext.md — current branch, last commit, pending tasks
2. .claude/memory/decisions.md — ADRs already decided (do not contradict)
3. This file — do not repeat known anti-patterns listed below

Project Structure

hooks/         102 Python hooks + shared libs (25 events in settings.json)
agents/        13 agent definitions + 3 squad teams (build / review / research)
skills/        158 skills — core/ (12) + extensions/ (146)
tests/         97 test files — pytest + bash smoke
rules/         9 markdown rules
mcp-profiles/  3 profiles: core / deploy / science
scripts/       inbox_review.py, populate_vault.py, skill-manager.sh
install.sh     One-command installer for other projects

Development Commands

python -m pytest tests/ -x -q --tb=short          # full suite
python -m pytest tests/<file>.py -v               # single file
ruff check hooks/ && mypy hooks/ --ignore-missing-imports
python -m pytest tests/ --cov=hooks --cov-report=term-missing
bash tests/smoke_skills.sh && bash tests/smoke_hooks.sh
cp -r skills/extensions/<name> ~/.claude/skills/extensions/<name>  # install skill globally

Coverage: ≥86% local/Windows · ≥65% CI/Linux

Skill Routing

What Agents CANNOT Do (without explicit user confirmation)

• git push — never without user approval
• git reset --hard, git rebase — destructive git ops
• Delete or disable tests — fix CODE, not tests
• Edit .env*, secrets, production config
• Launch simulations, ML training, or heavy computation autonomously
• Modify settings.json without update-config skill

New Skill Checklist

1. Create skills/extensions/<name>/SKILL.md with BSV header + YAML frontmatter
2. Register in skills/registry.yaml
3. Copy to ~/.claude/skills/extensions/<name>/
4. No tests required for skills (SKILL.md only)

Known Anti-Patterns — AVOID

• [×3] datetime.utcnow() mixed with timezone-aware datetimes → use datetime.now(timezone.utc)
• [×2] Coverage overclaim — README states metric without tool verification
• [×2] squash merge with 2+ commits → second commit lost; run git log --oneline on main after merge
• [×2] Forward-only workflow without integrity gates

Evidence Policy

• [VERIFIED] — confirmed with tool (Read/Bash/pytest)
• [INFERRED] — logical chain from verified facts
• [UNKNOWN] — no confirmation. Never fabricate metrics or test results.

Memory Protocol

Agents do NOT write to memory files directly — return results to orchestrator. Exception: tester agent may append one line to activeContext.md ## Test Status.