dev-process-toolkit

Implement a feature or fix end-to-end. Analyzes the request, builds in TDD order, runs gate checks, self-reviews with bounded loops, and reports for human approval before committing.

Bundle the Release Checklist + /docs --commit --full into one atomic, human-approved release commit. Reads specs/plan/M<N>.md, bumps the four release files, regenerates docs, prompts once for approval, commits on `y`, does not push.

Guide the user through writing or completing spec files (requirements, technical spec, testing spec, plan). Use after /setup to fill in specs before implementation, or to update existing specs.

Set up SDD/TDD development process for the current project. Creates CLAUDE.md, configures settings, and optionally creates spec files. Use when starting a new project or adding process to an existing one.

Socratic design session for greenfield features with open solution spaces. Clarifies goals one question at a time, proposes 2-3 approaches with tradeoffs, gets design approval, then feeds into /spec-write.

Run the project gate checks and report results. Use after completing any feature, before creating a PR, or to verify project health.

Capture a structured bug report — narrative + redacted curated context (and optional session transcript) — and publish to a secret GitHub gist for triage by the maintainer or self-debug via /brainstorm.

Manually archive user-selected FRs or milestones by `git mv` into `specs/frs/archive/` (and optionally `specs/plan/archive/`) with a diff approval gate. Accepts ULID, tracker ID/URL, or `M<N>`. Escape hatch for /implement Phase 4 auto-archival gaps.

Create a pull request with conventional format. Use when asked to create a PR, open a pull request, or push changes for review.

Drive `/smoke-test` against both trackers in parallel and aggregate the per-tracker findings files into one deduplicated report. Default `capture-only` mode honors `/smoke-test`'s "Capture, don't fix" rule unchanged. Opt-in `--auto-fix` mode walks the deduplicated high-severity findings list and dispatches `/dev-process-toolkit:spec-write` + `/dev-process-toolkit:implement` per finding, then re-iterates until termination. Project-local skill, not plugin.

Spawn a fresh Bun project under ../dpt-test-project-<tracker> and drive the dev-process-toolkit plugin's full skill chain (/setup → /spec-write → /implement → /gate-check → /spec-review → /simplify) via claude-st -p child sessions, capturing findings. Pre-release sanity check, not CI. Real Linear or Jira writes (per `--tracker`), ~10 min wall-clock. Two-terminal tandem runs (one per tracker) are supported.

Visually verify web UI using a rubber duck MCP with Chrome browser tools. Use after building or modifying web UI components.

Review implementation against specs to find deviations, missing features, or inconsistencies. Delegates to the `spec-review-audit` fork + `spec-reviewer` subagent.

Multi-agent TDD orchestrator. Runs RED → GREEN → REFACTOR → AUDIT for one FR via four forked subagents (test-writer / implementer / refactorer / spec-reviewer) with a strict tdd-result hand-off contract, bounded retries, and a deterministic halt path. Replaces the previous single-context /tdd.

Internal spec-review audit fork — invoked exclusively by /dev-process-toolkit:spec-review via context:fork pairing with the spec-reviewer subagent. Audits the FR's specs against shipped code (read-only) and emits a spec-review-result fenced block. Do not invoke directly.

Internal research fork — invoked exclusively by /brainstorm and /spec-write via context:fork pairing with the deps-researcher subagent. Returns topic-aware sibling-package retrieval as a ≤25-line block. Do not invoke directly.

Manage the sibling-package dependency manifest at `specs/deps.yaml`. Add, edit, delete, list, or sync entries that feed topic-aware retrieval into /brainstorm and /spec-write via the deps-researcher fork.

Internal TDD spec-review fork — invoked exclusively by /dev-process-toolkit:tdd via context:fork pairing with the tdd-spec-reviewer subagent, exactly once at the end of an FR after REFACTOR is GREEN. Audits the FR's ACs (read-only) and emits a tdd-spec-review-result fenced block. Do not invoke directly.

Internal research fork — invoked exclusively by /brainstorm and /spec-write via context:fork pairing with the spec-researcher subagent. Returns related-FR retrieval as a ≤25-line block. Do not invoke directly.

Internal TDD implementer fork — invoked exclusively by /dev-process-toolkit:tdd via context:fork pairing with the tdd-implementer subagent, once per AC. Implements the minimum code to turn one AC's failing test GREEN. Do not invoke directly.

Internal TDD refactorer fork — invoked exclusively by /dev-process-toolkit:tdd via context:fork pairing with the tdd-refactorer subagent, exactly once at end of FR after all ACs are GREEN. Do not invoke directly.

Internal TDD test-writer fork — invoked exclusively by /dev-process-toolkit:tdd via context:fork pairing with the tdd-test-writer subagent. Writes failing tests for the full AC list of one FR. Do not invoke directly.

Review changed code for reuse, quality, and efficiency, then fix any issues found. Use after completing a feature to clean up.

Generate or update project docs (requires /setup --docs first). One of --quick, --commit, --full. `--quick` stages a fragment per FR; `--commit` merges staged fragments with human approval; `--full` regenerates the canonical docs/ tree from scratch.

Structured debugging protocol for failed gate checks, failing tests, or unexpected behavior. Use when /gate-check or /tdd fails and the cause is unclear. Phases: Root Cause → Pattern Analysis → Hypothesis Testing → Implementation.