tsa-edit-then-verifylisted

The full edit-and-verify loop mandated by CLAUDE.md and docs/agent-tooling-gap-report.md:58. Pre-edit gate (safe_to_edit + baseline file_health) → LLM edits → post-edit verify (file_health diff + analyze_change_impact + scoped verification_command). Replaces "edit then run the whole pytest suite" (~5 min) with a scoped verification (~30-60s) that only escalates to full suite when risk remains. Use when: - About to make any non-trivial edit to a code file in tree-sitter-analyzer - User asks "implement X" / "fix this bug" / "refactor Y" / "add this feature" - You just finished editing and need to verify before committing - Any change touching tree_sitter_analyzer/** or tests/** Don't use when: - One-line typo in docs/comments (no behavior change → no verify needed) - Pure README/markdown edits - You're still in plan mode and haven't decided on the edit yet - File you're literally creating from scratch (no callers, no baseline) Workflow: 1 parallel batch pre-edit → edit → 1 parallel batch post-edit → run scoped
aimasteracc/tree-sitter-analyzer · ★ 36 · AI & Automation · score 78

Install: claude install-skill aimasteracc/tree-sitter-analyzer

# tsa-edit-then-verify — The full edit loop, scoped not full-suite > **Source of authority.** From `docs/agent-tooling-gap-report.md` line 58: > > > Every feature update must run the self-hosted workflow: `safe_to_edit` > > before risky edits, `file_health` on changed files, `change_impact` after > > edits, its reported `verification_command`, then the full default suite > > when risk remains. > > This skill is the **executable form** of that rule. `tsa-edit-safety` only > covers Phase A (the pre-edit gate). This skill covers A + B + C end-to-end > and is the canonical replacement for the global `verify` skill (which runs > the full ~5-minute suite every time). ## When to use vs. not **Use** when you're about to edit (or just edited) a real code file in this repo and the change is more than a 1-line cosmetic fix. The smaller the diff, the bigger the win — full pytest is ~5 min, scoped verification is typically 30-60s. **Don't use** for: - One-line typo in a doc/comment (no Python AST impact) - Pure markdown / README edits (no symbol-level risk) - File you're creating from scratch with no callers yet - Read-only exploration ## Why this exists (vs. `verify` skill) | Skill | What it runs | Time | When right | |-----------------------------|---------------------------------------|---------|------------| | `verify` (global) | `uv run pytest -q` (15k tests) | ~5 min | Pre-commit gate, large refactor | | **`ts