adr
Write an Architecture Decision Record (ADR) for a feature — Context / Decision / Status / Consequences / Alternatives, filed as…
Quick pre-commit checks — lint:fix -> test
$ npx -y skills add sd0xdev/sd0x-dev-flow --skill precommit-fast --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/precommit-fastContext preview
The summary Claude sees to decide when to auto-load this skill.
Quick pre-commit checks — lint:fix -> test
name: precommit-fast description: "Quick pre-commit checks — lint:fix -> test" allowed-tools: Bash(node:*), Bash(pnpm:*), Bash(yarn:*), Bash(npm:*), Bash(npx:*), Bash(python*:*), Bash(pytest:*), Bash(ruff:*), Bash(mypy:*), Bash(cargo:*), Bash(go:*), Bash(golangci-lint:*), Bash(./gradlew:*), Bash(mvn:*), Bash(bundle:*), Bash(rubocop:*), Bash(rspec:*), Bash(git:*), Read, Grep, Glob
| Step | Goal | Safety | Skip if Missing | |------|------|--------|----------------| | lint-fix | Auto-fix code style issues | read-write | yes | | test-unit | Run fast test suite | read-only | yes |
**Failure behavior**: continue-all (run all steps, report all results)
Run quick pre-commit checks: **lint:fix -> test** (no build step)
Use Glob to check if `.claude/scripts/precommit-runner.js` exists in the project root.
1. **Manifest gate**: Use Glob to check if any known manifest exists (`package.json`, `pyproject.toml`, `Cargo.toml`, `go.mod`, `build.gradle`, `build.gradle.kts`, `pom.xml`, `Gemfile`). None → skip, fall through to Step 2. (The runner orchestrates every ecosystem in the Step 2 table, so a non-Node manifest justifies the install just as well.) 2. **Locate plugin scripts**: 3-level Glob fallback (short-circuit on first match):
3. **Plugin not found** → fall through to Step 2. 4. **Plugin found** → copy runner + lib/utils.js + review-state.js + lib/tree-digest.js (skip on conflict) → run. The checker pair is the runner's self-note dependency — without it the runner still runs every check and reports; only the advisory note is skipped, loudly.
Detect the project ecosystem to run steps manually.
| Manifest | Ecosystem | Lint-fix | Test | |----------|-----------|----------|------| | `package.json` | Node.js | `{pm} lint:fix` | `{pm} test:fast` / `test:unit` / `test` | | `pyproject.toml` | Python | `ruff check --fix .` | `pytest tests/unit/` | | `Cargo.toml` | Rust | `cargo clippy --fix` | `cargo test` | | `go.mod` | Go | `golangci-lint run --fix` | `go test ./...` | | `build.gradle` / `build.gradle.kts` | Java | `./gradlew spotlessApply` | `./gradlew test` | | `pom.xml` | Java (Maven) | `mvn spotless:apply` | `mvn test` | | `Gemfile` | Ruby | `bundle exec rubocop -a` | `bundle exec rspec` |
> **How the runner executes this table**: same semantics as `skills/precommit/SKILL.md` § Ecosystem detection — a missing required tool is `unavailable` and blocks ✅ PASS; repo-declared capabilities (spotless / rubocop / rspec) skip only on **definitive tool-native absence evidence** (lockfile membership, the tool's own not-found diagnostic); an ambiguous or timed-out probe is `unavailable`, never a skip.
After lint:fix completes, run `git diff --name-only` to capture auto-fixed files.
**After a conclusive fallback run, self-note the outcome** — `note precommit pass` when every executed check passed, `note precommit fail` when the checks ran and at least one failed (the `rounds` count stays path-independent):
CHECKER=".claude/scripts/review-state.js"; [ -f "$CHECKER" ] || CHECKER="scripts/review-state.js" node "$CHECKER" note precommit pass # or fail
An **inconclusive** run (checks could not execute) notes nothing — the slot stays untouched, like the runner's `⚠️ NO CHECKS RUN`. The note is advisory; a failed note never fails the run.
## Precommit (fast) ## Results | Step | Status | Notes | |------|--------|-------| | lint:fix | ✅/❌/⏭️/⛔ | ⏭️ = repo opted out (skip); ⛔ = required tool unavailable (blocks PASS) | | test | ✅/❌/⏭️/⛔ | same legend | ## Changed Files (after lint:fix) - <files or "(none)"> ## Overall: ✅ PASS / ❌ FAIL / ⚠️ NO CHECKS RUN (no project validation executed AND no policy step failed; fall through to Step 2, needs human if nothing runnable exists) ## Checklist - [ ] All available checks pass - [ ] git status reviewed
Language: English | 繁體中文 | 简体中文 | 日本語 | 한국어 | Español The harness layer for Claude Code. Let the model choose the path. Keep "done" verifiable. Full control plane on Claude Code. Skills-only distribution for Codex CLI and other compatible agents.
Repo: sd0xdev/sd0x-dev-flow
Write an Architecture Decision Record (ADR) for a feature — Context / Decision / Status / Consequences / Alternatives, filed as…
Architecture design and documentation. Produces 3-architecture.md with component diagrams, data flow, integration points, and architecture decisions. Reads…
Context-aware Q&A with auto context gathering. Use when: user has a quick question about codebase, git history, rules, docs, or skills during development. Not…
Industry best practices conformance audit with mandatory adversarial debate. Produces audit artifact: verdict (OK/WARN/FAIL) + gap roadmap + debate proof. Use…
Bug fix workflow. Use when: fixing bugs, resolving issues, regression fixes. Not for: new features (use feature-dev), understanding code (use code-explore).…
Bump package and plugin version in sync. Updates package.json, .claude-plugin/plugin.json, and install-state manifest to the same version. Use when: user says…