Independently test semantics, keyboard flow, focus, contrast, reduced motion, zoom, touch targets, and error messaging.
Independently review frozen artifacts and report reproducible evidence-backed defects before seeing builder reasoning.
Implement APIs, services, jobs, validation, authorization, reliability, idempotency, and targeted tests.
Create original brand, logo, app icon, favicon, and coherent production icon rules with provenance.
Maintain the task graph, deadlines, dependencies, assignments, checkpoints, and concise operating summaries. Use for orchestration state.
Run the A2Z FORGE design research tournament and establish the design constitution. Use before building a new interface or when a product direction must change.
Produce the founder debrief and technical handoff. Use when a milestone or product needs to be understandable and maintainable by its owner.
Create or update the living product goal and blueprint. Use when scope, north star, positioning, non-goals, success measures, or product decisions change.
Run A2Z FORGE rapid intake or the adaptive founder interview. Use when an idea lacks material founder, problem, product, business, design, security, or delivery constraints.
Run bounded persistent orchestration until a milestone, blocker, budget boundary, deadline, or user stop. Use for autonomous continuation with explicit limits.
Executes bash commands
Hook triggers when Bash tool is used
Modifies files
Hook triggers on file write and edit operations
Uses power tools
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.
agent_limitMaximum concurrent workers used by A2Z FORGE.
${user_config.agent_limit}permission_modeDefault operating boundary: observe, build, connected-build, or autopilot.
${user_config.permission_mode}reserve_percentPercentage of estimated capacity held for integration and critical fixes.
${user_config.reserve_percent}Uses Bash, Write, or Edit tools
Uses Bash, Write, or Edit tools
Idea in. Production out.
A2Z FORGE is an autonomous product foundry for Claude Code and compatible coding agents. It coordinates research, design, implementation, independent critique, testing, recovery, demonstration, and handoff through one evidence-driven workflow.
This repository is under active construction. The recoverable orchestration foundation is implemented first. Production-readiness claims will appear only beside reproducible evidence.
A2Z FORGE does not guarantee product-market fit, revenue, perfect output, exact provider-limit avoidance, or that simulated review equals customer validation.
npm install
npm run build
node dist/cli/index.js init --path /tmp/forge-demo --idea "Build a focused study workspace"
node dist/cli/index.js status --path /tmp/forge-demo
node dist/cli/index.js checkpoint --path /tmp/forge-demo
node dist/cli/index.js recover --path /tmp/forge-demo
./installers/install.sh
Restart Claude Code and invoke /a2z-forge:forge <idea>. Plugin skills are namespaced by Claude Code; see installation details.
See IMPLEMENTATION_PLAN.md and docs/ENVIRONMENT_REPORT.md. The Claude plugin, full adapter set, dashboard, and reference application are not yet claimed complete.
npx claudepluginhub jumpsy/a2z-forge --plugin a2z-forgeTurns Claude Code into a senior+ engineering peer that ships production-grade software: architecture, TDD, strict types, security, performance, observability, and taste. No hand-holding. The inverse of noob-skill.
Turns Claude Code into a patient, premium AI-app builder for people who have never coded. Ship your first secure, good-looking AI app one step at a time.
Evidence-gated AI coding workflow: scan → analyze → plan → TDD → execute → fix → verify → review, powered by Codebase Memory MCP >= 0.9.0 with optional Serena LSP intelligence. Includes blast-radius planning, test/cycle gates, independent review, and Windows Git Bash hook auto-resolution.
Lazy senior dev mode. Forces the simplest, shortest solution that actually works: YAGNI, stdlib first, no unrequested abstractions.
Consult multiple AI coding agents (Gemini, OpenAI, Grok, Perplexity, plus codex, antigravity, and grok CLIs when installed) to get diverse perspectives on coding problems
Production-grade engineering skills for AI coding agents — covering the full software development lifecycle from spec to ship.
Feature development with code-architect/explorer/reviewer agents, CLAUDE.md audit and session learnings, and Agent Skills creation with eval benchmarking from Anthropic.
Comprehensive skill pack with 66 specialized skills for full-stack developers: 12 language experts (Python, TypeScript, Go, Rust, C++, Swift, Kotlin, C#, PHP, Java, SQL, JavaScript), 10 backend frameworks, 6 frontend/mobile, plus infrastructure, DevOps, security, and testing. Features progressive disclosure architecture for 50% faster loading.