Use when asked to improve the AIDP plugin itself by finding exactly one small maintainer-reviewable bug, docs ambiguity, confusing error, test gap, or contributor-experience papercut.
Use when running, reviewing, or auditing software work with AIDP v2 typed artifacts, bounded context, proof commands, and completion audits.
AIDP v2 is a typed, agentic kernel for AI-assisted development with bounded context and direct proof. It records Goal, context, constraints, and done-when in versioned artifacts, then requires line spans, command logs, and typed artifacts before a run can finish.
aidp "<objective>" --run-id <run-id> --proof-command "<command>"
aidp agent <run-id>
aidp check <run-id>
aidp prove <run-id>
aidp finish <run-id>
The bare objective command creates .aidp/runs/<run-id>/contract.json, context.json, plan.json, events.jsonl, and a compact work.md. The agent handoff prints an executable prompt that references the packet instead of embedding source bodies.
| Command | Purpose |
|---|---|
aidp doctor | Detect project languages, package manager, and likely proof commands. |
aidp "<objective>" | Start a run and write a compact worker packet. |
aidp plan "<objective>" | Create typed run artifacts without writing a worker packet. |
aidp work <run-id> | Regenerate .aidp/runs/<run-id>/work.md from artifacts. |
aidp agent <run-id> | Print and persist an executable agent prompt. |
aidp check <run-id> | Validate typed artifacts and schema versions. |
aidp prove <run-id> | Run proof commands and store command evidence. |
aidp finish <run-id> | Block unless every done-when requirement has direct evidence. |
aidp adapter codex | Render compact Codex AGENTS.md guidance for the v2 flow. |
aidp self-test | Run hermetic docs-only and code/test scenarios against the v2 lifecycle. |
aidp cutover | Print a non-destructive audit for retired surfaces before any deletion patch. |
Every run lives under .aidp/runs/<run-id>/.
| Artifact | Role |
|---|---|
contract.json | Objective, constraints, risk, unsafe actions, and done-when requirements. |
context.json | Bounded source pointers, detected commands, citations, and omitted context. |
plan.json | Expected tasks, test-first notes, and proof commands. |
work.md | Compact worker packet for an implementation agent. |
agent.md | Executable handoff prompt generated from the packet. |
evidence.json | Command evidence, changed files, review results, risks, and verdict. |
events.jsonl | Append-only lifecycle events. |
Schemas for the typed artifacts live in docs/v2/.
This repo is packaged for Codex and Claude Code. Both packages reuse the same skills/ directory and the same aidp CLI.
| Host | Manifest | Local test/install path |
|---|---|---|
| Codex | .codex-plugin/plugin.json | Use this repository as the Codex plugin root. |
| Claude Code | .claude-plugin/plugin.json | Run claude --plugin-dir . from this repository root. |
Included skills:
| Skill | Purpose |
|---|---|
aidp-v2 | Run normal AIDP v2 work through typed artifacts, bounded context, proof commands, and completion audits. |
aidp-maintainer-improvement | Find exactly one small maintainer-reviewable improvement, make the minimum patch, verify it, and stop. |
Claude Code exposes the skills as namespaced plugin skills such as /ai-dev-pipeline:aidp-v2 and /ai-dev-pipeline:aidp-maintainer-improvement.
For changes to AIDP itself, run:
uv run pytest -q
uv run aidp self-test
uv run aidp cutover
uv run pytest -q covers the v2 package, command lifecycle, plugin surface, and completion audit. aidp self-test proves both docs-only and code/test scenarios through agent, check, prove, and finish. aidp cutover should report a clean retired-surface audit before release.
The ground-up redesign is documented in docs/redesign/2026-06-12-ground-up-ai-dev-pipeline.md.
Own this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimOwn this plugin?
Verify ownership to unlock analytics, metadata editing, and a verified badge. GitHub access is read-only (username + org membership).
Sign in to claimBased on adoption, maintenance, documentation, and repository signals. Not a security audit or endorsement.
npx claudepluginhub 0x7067/ai-dev-pipeline --plugin ai-dev-pipelineReplaces str_replace-style editing in Claude Code with line-anchored hashline patches (via an MCP server) and blocks the built-in Edit/Write/NotebookEdit tools.
Elite AI development framework: reference-first design, agent orchestration, automated quality gates, and battle-tested engineering workflows
Persona-driven AI development team: orchestrator, team agents, review agents, skills, slash commands, and advisory hooks for Claude Code
AI-powered development workflow automation - Phase-based planning, implementation orchestration, preflight code quality checks with security scanning, ship-it workflow, and development principles generator for CLAUDE.md
AI development loop — orchestrator distributes tasks to headless workers, independent auditor verifies, structural enforcement auto-blocks downstream on upstream failure. Full-cycle validated with 10-scenario test harness.
SDLC enforcement for AI agents — TDD, planning, self-review, CI shepherd
Make AI coding agents follow a repeatable engineering workflow with memory, verification, skills, and multi-agent setup