Spaces — Development skill for Claude Code

Spaces

**An open-source, agent-driven SDLC orchestrator for software development teams.**

Spaces runs an AI-driven Software Development Life Cycle — `specify → plan → tasks → implement → verify` — across a fleet of specialized agents, with a web UI to inspect every step, human-in-the-loop review gates, and app-wide OAuth integrations for GitHub, Jira, Confluence, and Slack.

Think of it as a project board where every card is powered by a persistent agent that knows the codebase, your team's conventions, and the artifacts of every previous stage.

Built on the [AIDLC framework](https://github.com/awslabs/aidlc-workflows) (AI-Driven Development Life Cycle) and the [Pi Coding Agent SDK](https://www.npmjs.com/package/@earendil-works/pi-coding-agent).

**Tags:** `agentic-workflows` · `aidlc` · `sdlc-automation` · `ai-development` · `llm-orchestration` · `pipeline-orchestrator` · `spec-kit` · `claude` · `developer-tools`

---

Screenshots

The board

Every Space appears as a card on a kanban board that auto-derives its lane (Backlog → Initialized → Specified → Planned → Tasked → Implementing → Done) from the artifacts each project has produced. The **Integrations chip** in the hero shows connection status for GitHub, Jira, Confluence, and Slack at a glance.


Project detail — pipeline running

Open a card and you get the AI agent output live, plus tabs for specs, tests, implementation, QA, chat, context, memory, and promotions. The orchestrator panel on the right controls autonomous mode, per-project concurrency, and the Fast/Balanced/Quality speed mode fed to the model router.


Generated artifacts, browsable in-app

Every stage produces markdown artifacts (`spec.md`, `plan.md`, `tasks.md`, `test-plan.md`, `verification-report.md`, etc.) that render inline in the app. Below