By zstevovich
Agent Pipeline Development — enforced multi-agent pipelines with mechanical guardrails
Use when verifying that APD is correctly configured in the current project — qualitative deep audit of agents, hooks, CLAUDE.md, pipeline state, MCP wiring, and guardrails. Goes further than verify-apd. Triggers on "audit APD", "review setup", "is APD configured", "verify framework", "check APD", "APD health", "is everything wired", after any major framework upgrade or version bump.
Use BEFORE writing the APD spec-card.md whenever the task is vague, broad, ambiguous, or has multiple reasonable approaches. Asks ONE question at a time, presents 2-3 options when there are real choices, converges on a design the user explicitly approves. Triggers on "improve X", "what should we", "thinking about", "options", "not sure", "maybe", "vague", "broad", "redesign", any spec-card.md with unclear scope or fewer than 3 R-criteria. For the pipeline gate contract (plan format, rationale format, BLOCKs) load /apd-pipeline-guide — that one is mandatory on every task; this one is for unclear scope.
MANDATORY before re-dispatching the builder after any APD verifier failure or critical reviewer finding. Use on any bug, test failure, build failure, error, crash, regression, or unexpected behavior during the APD pipeline. Four-phase root-cause analysis — investigate, pattern-match, hypothesize + controlled test, fix with TDD. Triggers on "bug", "fail", "failing", "broken", "doesn't work", "error", "crash", "regression", "verifier blocked", "reviewer rejected".
MANDATORY after every successful APD pipeline commit. Use as soon as the pipeline reports commit done — verifies tests from a clean state, shows the pipeline report, presents four options (push, push+PR, keep local, discard), executes the user's choice. Triggers on "after commit", "push", "PR", "pull request", "deploy", "next", "pipeline done", "shipped", commit hash present in pipeline state.
Use when the project has GitHub Projects configured (GITHUB_PROJECTS_URL in CLAUDE.md) and APD pipeline tasks need to sync with the board. Creates issues for specs, moves cards through columns (todo → in-progress → review → done), closes on commit. Triggers on "GitHub Projects", "issue", "board", "card", "sync", "kanban", "project board", any APD pipeline phase transition when a board URL is configured.
Matches all tools
Hooks run on every tool call, not just specific ones
Executes bash commands
Hook triggers when Bash tool is used
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.
This plugin requires configuration values that are prompted when the plugin is enabled. Sensitive values are stored in your system keychain.
stackBackend stack: nodejs, php, python, dotnet, go, java
${user_config.stack}author_nameAuthor name for git commits (no AI signatures)
${user_config.author_name}project_nameYour project name (e.g., MyCRM, TaskFlow)
${user_config.project_name}Modifies files
Hook triggers on file write and edit operations
Modifies files
Hook triggers on file write and edit operations
Enforced multi-agent pipelines with mechanical guardrails for AI-assisted software development.
Works with Claude Code and OpenAI Codex.
Getting Started · Interactive Demo · Pipeline Runs · Changelog
v6.16.1 · MIT · macOS + Linux

APD enforces a disciplined workflow where specialised agents build code through a gated pipeline. Every step is mechanically enforced — hooks block violations, not documentation.
Spec → Builder → Reviewer → [Adversarial] → Verifier → Commit
Claude Code:
/plugin marketplace add zstevovich/claude-apd # one time
/plugin install claude-apd@zstevovich-plugins # per project
/apd-setup # configure
bash .claude/bin/apd verify # check setup
Codex (OpenAI) — preferred, via plugin marketplace (validated on Codex 0.124):
# Prerequisites: brew install uv jq (codex_hooks is stable on 0.124, no feature flag needed)
codex plugin marketplace add zstevovich/claude-apd # one time, pulls latest main
printf '\n[plugins."apd@codex-apd"]\nenabled = true\n' >> ~/.codex/config.toml # enable
cd <your-project> && codex # open TUI in target project
# On first user prompt in the fresh session, APD auto-scaffolds
# .codex/ + .apd/ + AGENTS.md and MCP tools become available
# (apd_ping, apd_doctor, apd_guard_write, ...).
Codex — alternative, direct-drop (no marketplace):
git clone https://github.com/zstevovich/claude-apd ~/apd && \
export PATH="$HOME/apd/bin:$PATH" # live source at ~/apd
apd cdx init # scaffold .codex/ + .apd/ + AGENTS.md
apd cdx agents add code-reviewer # populate agents
apd cdx agents add backend-builder src/ config/
apd cdx skills install # symlink phase skills into ~/.codex/skills/
apd cdx doctor # audit setup
codex # start session; "run apd_ping"
Claude Code:
/plugin update claude-apd@zstevovich-plugins # pulls latest from main
# Restart the CC session — skill slash commands re-register on session start
Codex — marketplace install (Git source):
codex plugin marketplace upgrade codex-apd # pulls latest main, replaces cache atomically
# Open a new TUI session — first user prompt picks up the new version
To pin the upgrade to a specific tag or branch (e.g. for pre-release testing):
codex plugin marketplace remove codex-apd
codex plugin marketplace add zstevovich/[email protected] # or @<branch> for a feature branch
Codex — direct-drop install:
bash .codex/bin/apd update # git pull --ff-only + idempotent re-init
# --check-only # dry-run, don't pull or write
# --skip-pull # only re-run install-codex-config locally
Codex SessionStart quirk (upstream). Per openai/codex#15269, Codex 0.124 fires the TUI
SessionStarthook when the user submits the first prompt, not at banner display. Practically: gap-analysis runs on first turn, not at open. MCP tools are available immediately.
See Getting Started for both walkthroughs.
| Role | Model | Effort | Responsibility |
|---|---|---|---|
| Orchestrator | opus | max | Coordinates pipeline, writes spec, dispatches agents, commits |
| Builder | sonnet | xhigh | Implements code per spec, scoped to specific files |
| Reviewer | opus | max | Finds bugs, security issues, edge cases (read-only) |
| Adversarial Reviewer | sonnet | max | Context-free review — no spec knowledge, fresh perspective |
| Verifier | — | — | Script: build + test + spec traceability check |
Model and effort per agent are per-project settings — and since v6.16 they are switchable as named profiles.
The model raises the average. APD guarantees the floor. The profile chooses what the average costs.
npx claudepluginhub zstevovich/claude-apd --plugin claude-apdMulti-agent /workflow development pipeline (planner → plan-review → coder → code-review) with typed handoff contracts, lifecycle hooks, and MCP servers.
Long-running agent harness with 5-layer memory architecture, GitHub integration, autonomous batch processing, Agent Teams with ATDD, 9 hooks (safety, quality gates, team coordination), and 6 Agent Skills
11 agents, 35 skills, 18 commands, 9 hooks — spec-driven multi-agent orchestration for Claude Code, with optional cross-device semantic memory.
Complete project development toolkit: 23 agents, 23 slash commands, 29 lifecycle hooks, and 69 reusable skills for Claude Code workflows
Portable Development System — AI-assisted development methodology with skills for consistency and agents for scale.
HarnessFlow — From idea to shipped product: high-quality engineering workflows for AI agents. Spec-anchored SDD, gated TDD, evidence-based routing, independent reviews, and formal closeout.