claude

atdd-pipeline

Multi-agent ATDD pipeline: impact map -> Gherkin -> GitHub Issues -> RED/GREEN with parallel reviewers and autonomous sub-PR merging. Two human gates: spec, and the final PR.

0 on GitHubby Mathieu Pitonv0.3.0MITupdated 1 month agoworkflow
/

Plugin atdd-pipeline from marketplace agentic-atdd in mpiton/agentic-atdd. Commands read from the repo's manifest — 1 month ago.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (16)

  • apply-pr-feedback

    Fetch bot and human review comments on the current PR, apply every actionable change in scope, commit and push. Used by pr-auto-merge inside the ATDD pipeline. Can also be invoked standalone via /apply-pr-feedback or /apply-pr-feedback <pr-number>.

  • green-cycle

    ATDD GREEN phase. Generate the minimal implementation that makes the RED test pass, run review-architecture and review-intent in parallel, auto-correct up to 2 times, then escalate. Use after red-cycle on the same scenario branch, or when the user types /green <issue-number>.

  • pr-auto-merge

    Watch a sub-PR through bot review + CI, run /apply-pr-feedback on actionable feedback, then auto-merge into its base branch. Refuses to act when base is the trunk (main/master). Use after green-cycle opens a draft PR, or when the user types /auto-merge <pr-number>.

  • red-cycle

    ATDD RED phase. Generate a failing acceptance test at the right level from a scenario sub-issue, run review-fidelity, auto-correct up to 2 times, escalate to human comment on failure. Use when the user types /red <issue-number> or the orchestrator iterates scenarios.

  • review-architecture

    Read-only review of an implementation diff for structural quality - placement, naming, project conventions, and domain responsibilities. Returns OK or REGENERATE. Invoked in parallel with review-intent by green-cycle, also callable standalone on a PR or diff.

  • review-fidelity

    Read-only review verifying that an acceptance test mirrors its Gherkin scenario in structure, semantics, and intent. Returns a verdict OK or REGENERATE. Invoked by red-cycle, also callable standalone on any test/scenario pair.

  • review-intent

    Read-only review of an implementation diff for intent conformance - code matches the scenario, no over-engineering, no hidden side effects. Returns OK or REGENERATE. Invoked in parallel with review-architecture by green-cycle.

  • verify-acceptance

    ATDD Verify stage. Checks that the integrated user story works the way a human would check it, blind to the acceptance tests - drives the real UI for @ui scenarios, the real HTTP/CLI interface for @e2e, a throwaway script for @use-case, and falls back to a code trace only when the app cannot run. Al

  • atdd-run

    End-to-end ATDD pipeline. Use when the user wants to take a user story from impact mapping all the way to merged code, or types /atdd-run.

  • setup-atdd-pipeline

    First-run configuration for the atdd-pipeline plugin in the current project. Picks an issue tracker, spec base path, commit style, and writes .atdd-pipeline.json at the repo root. Idempotent. Use after installing the plugin in a new repo.

  • from-issue

    Import an existing GitHub issue into the ATDD pipeline. Parses the issue body into a context.md, interviews the user to fill gaps, tags the issue, and hands off to spec-generate. Use when the user has an existing issue tracker entry they want to drive through the pipeline.

  • from-linear

    Import a Linear ticket into the ATDD pipeline. Fetches the card via Linear MCP, CLI, or GraphQL API (first one available), parses the description into a context.md, interviews the user to fill gaps, and hands off to spec-generate. Use when the story lives in Linear (ticket like ENG-123).

  • impact-map

    Capture a user story, actor, action, goal, and business rules from the user. Use when the user wants to start a new feature and has not yet provided structured business context, or types /impact-map.

  • spec-generate

    Generate Gherkin acceptance scenarios from a context.md, with mandatory interview for ambiguities, concrete data, and test-level tags. Use after impact-map, or when the user provides a user story and rules and asks for scenarios.

  • spec-review

    Read-only review of generated Gherkin scenarios across four axes (branches, coherence, gaps, triangulation). Produces a verdict OK or REGENERATE. Use after spec-generate.

  • to-issues-atdd

    Idempotent sync of ATDD spec artifacts to GitHub Issues. Creates parent / user story / one sub-issue per scenario, with test-level labels and milestone. Re-runnable without duplication. Use after spec-review verdict is OK.

Subagents (3)

  • atdd-merge

    Serialized merge of ONE scenario sub-PR into the integration branch. Dispatched by atdd-run after each scenario (sequential mode) or by the atdd-stage3 workflow inside a single-lane mutex, so only one merge runs at a time. Rebases the branch onto the live integration tip, watches CI and bot review,

  • atdd-scenario

    Runs the ATDD RED then GREEN cycle for ONE scenario sub-issue in its own context. Dispatched by atdd-run - one at a time in the main checkout (sequential mode), or one isolated worktree per scenario in parallel (atdd-stage3 workflow). Writes the failing test, makes it pass minimally, runs the review

  • atdd-verify

    Verifies the integrated user story the way a human QA would, blind to the acceptance tests - real browser for @ui scenarios, the real HTTP/CLI interface for @e2e, a throwaway script for @use-case. Dispatched by atdd-run at Stage 3.5 so screenshots, DOM snapshots and app logs stay out of the orchestr

Hooks (1)

  • PreToolUse

    "${CLAUDE_PLUGIN_ROOT}"/hooks/guard-merge.sh

View source on GitHub

Similar plugins

Matt Pocock's agent skills for real engineering (grilling, spec/ticket flows, TDD, code review, domain modelling and more).

39 skills

mattpocock/skills 323,108

superpowers

claudecodex

Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques

15 skills · hooks

obra/superpowers 316,836

Backend API design, GraphQL architecture, workflow orchestration with Temporal, and test-driven backend development

9 skills · 1 command · 8 agents

wshobson/agents 42,461

beads

claudecodex

AI-supervised issue tracker for coding workflows

1 skill · 1 agent · hooks

gastownhall/beads 28,698

MiniMax AI skills library for frontend, fullstack, Android, iOS, shader, GIF sticker, document, presentation, spreadsheet, and multimodal media workflows

18 skills

MiniMax-AI/skills 13,828

worktrunk

claudecodex

Worktrunk is a CLI for Git worktree management, designed for parallel AI agent workflows. This plugin provides configuration guidance (LLM commit messages, project hooks, worktree paths), automatic activity tracking (🤖/💬 indicators in `wt list` showing active Claude sessions), and a `/wt-diff` pane showing each worktree's changes.

2 skills · hooks

max-sixty/worktrunk 12,481

Maintain atdd-pipeline?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex pluginslisted