{"id":45969657,"url":"https://github.com/arch-team/devpace","last_synced_at":"2026-04-02T00:45:32.969Z","repository":{"id":340307211,"uuid":"1162612957","full_name":"arch-team/devpace","owner":"arch-team","description":"BizDevOps rhythm manager for Claude Code — trace business goals to code changes with 18 skills","archived":false,"fork":false,"pushed_at":"2026-03-22T16:53:22.000Z","size":3243,"stargazers_count":42,"open_issues_count":1,"forks_count":2,"subscribers_count":0,"default_branch":"main","last_synced_at":"2026-03-22T23:51:32.228Z","etag":null,"topics":["ai-coding","ai-development","bizdevops","change-management","claude-code","claude-code-plugin","devops","plugin","project-management","quality-gates"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"mit","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/arch-team.png","metadata":{"files":{"readme":"README.md","changelog":"CHANGELOG.md","contributing":"CONTRIBUTING.md","funding":null,"license":"LICENSE","code_of_conduct":"CODE_OF_CONDUCT.md","threat_model":null,"audit":null,"citation":null,"codeowners":".github/CODEOWNERS","security":"SECURITY.md","support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null,"zenodo":null,"notice":null,"maintainers":null,"copyright":null,"agents":null,"dco":null,"cla":null}},"created_at":"2026-02-20T13:35:09.000Z","updated_at":"2026-03-22T16:07:32.000Z","dependencies_parsed_at":null,"dependency_job_id":null,"html_url":"https://github.com/arch-team/devpace","commit_stats":null,"previous_names":["arch-team/devpace"],"tags_count":17,"template":false,"template_full_name":null,"purl":"pkg:github/arch-team/devpace","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arch-team%2Fdevpace","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arch-team%2Fdevpace/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arch-team%2Fdevpace/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arch-team%2Fdevpace/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/arch-team","download_url":"https://codeload.github.com/arch-team/devpace/tar.gz/refs/heads/main","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/arch-team%2Fdevpace/sbom","scorecard":null,"host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":286080680,"owners_count":31293517,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2026-04-01T21:15:39.731Z","status":"ssl_error","status_checked_at":"2026-04-01T21:15:34.046Z","response_time":53,"last_error":"SSL_read: unexpected eof while reading","robots_txt_status":"success","robots_txt_updated_at":"2025-07-24T06:49:26.215Z","robots_txt_url":"https://github.com/robots.txt","online":false,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["ai-coding","ai-development","bizdevops","change-management","claude-code","claude-code-plugin","devops","plugin","project-management","quality-gates"],"created_at":"2026-02-28T15:04:11.091Z","updated_at":"2026-04-02T00:45:32.964Z","avatar_url":"https://github.com/arch-team.png","language":"Python","readme":"🌐 [中文版](README_zh.md) | English\n\n# devpace\n\nGive your Claude Code projects a steady development pace — requirements change, rhythm stays.\n\n\u003e A development harness for Claude Code — rules, schemas, gates, and feedback loops that keep AI-assisted development traceable and measurable.\n\n![version](https://img.shields.io/github/v/release/arch-team/devpace?label=version) ![license](https://img.shields.io/badge/license-MIT-green) ![type](https://img.shields.io/badge/Claude%20Code-Plugin-purple)\n\n## Why devpace\n\nWhen using Claude Code for product development:\n\n| Problem | Without devpace | With devpace |\n|---------|----------------|-------------|\n| Requirements change = chaos | \"Add a feature\" triggers cascading confusion, nobody knows the blast radius | Impact analysis + orderly adjustment, Claude won't change the plan on its own |\n| Inconsistent quality | Claude sometimes skips tests, forgets checks | Auto-checks + human approval, quality gates cannot be bypassed |\n| Work drifts from goals | Technical work disconnects from business goals, lots done but unclear value | Goal-to-code traceability at all times |\n| Re-explain everything each session | Manual approach needs **8** user corrections (3 interruption test) | Auto-restores context, **0** corrections |\n\n\u003e [See full walkthrough: from init to done](examples/todo-app-walkthrough.md)\n\n## 30-Second Experience\n\n```\n/pace-init                 ← Initialize (once)\n\"Help me implement user auth\" ← Claude auto-tracks tasks, writes code, runs tests, checks quality\n\"approve\" or \"reject\"      ← You decide whether to merge\n\"add an export feature\"    ← Claude analyzes impact, adjusts plan, waits for your confirmation\n\"pause auth for now\"       ← Impact analysis → pause (preserving work) → adjust plan\n```\n\nNext session, Claude reports: \"Last time we stopped at auth module, continue?\" — zero manual re-explanation.\n\n## Full Development Lifecycle\n\n```\nOpportunity ──→ Epic ──→ Requirements ──→ Features ──→ Code Changes ──→ Quality ──→ Ship\n        │          │           │              │              │              │           │\n     pace-biz   pace-biz   pace-biz       pace-plan      pace-dev     pace-review  pace-release\n                                          pace-change                              pace-feedback\n```\n\nRequirements can change anytime — `/pace-change` auto-analyzes impact, adjusts the plan, and waits for your confirmation.\n\nAfter each cycle, `/pace-retro` shows quality metrics and improvement trends.\n\n## How It Works\n\ndevpace builds a **goal-to-code traceability chain** in your project:\n\n1. **Goal alignment** — Every code change links back to a business goal. No work without purpose.\n2. **Auto quality gates** — Claude auto-checks code quality and requirement consistency, self-repairs on failure. Human approval cannot be skipped.\n3. **Change is normal** — Requirements changed? Auto impact analysis, orderly adjustment, existing work preserved.\n4. **Interrupt-proof** — Session broke? Auto-resume next time. All state in `.devpace/` as plain Markdown.\n\nUnder the hood: a Claude Code Plugin using Rules (behavioral guidelines) + Skills (`/pace-*` commands) + Hooks (auto-triggers at critical moments).\n\n## Installation\n\n\u003e **Prerequisite**: [Claude Code CLI](https://claude.ai/code) must be installed.\n\n### Marketplace Install (Recommended)\n\n```bash\n# Step 1: Register marketplace (one-time)\n/plugin marketplace add arch-team/devpace-marketplace\n\n# Step 2: Install\n/plugin install devpace@devpace\n```\n\n### From Source\n\n```bash\ngit clone https://github.com/arch-team/devpace.git\nclaude --plugin-dir /path/to/devpace\n```\n\n### Verify Installation\n\nAfter installing, type `/pace-` in Claude Code. If devpace is loaded, you'll see auto-complete suggestions for `/pace-init`, `/pace-dev`, `/pace-status`, etc.\n\n## Commands\n\nMost of the time you don't need commands — saying \"help me implement X\" equals `/pace-dev`, \"where are we\" equals `/pace-status`.\n\n### Start Here (5 commands cover 90% of daily work)\n\n| Command | Purpose |\n|---------|---------|\n| `/pace-init` | First-time setup (once) |\n| `/pace-dev` | Start coding — or just say \"help me implement X\" |\n| `/pace-status` | Check progress — or say \"where are we\" |\n| `/pace-review` | Approve changes |\n| `/pace-next` | Not sure what's next? Get AI recommendations |\n\n### More Commands\n\nAs your project grows, devpace offers change management, iteration planning, business alignment, release orchestration, and more. See [User Guide](docs/user-guide.md) for all 19 commands.\n\nQuick reference: `/pace-change` (requirements changed) · `/pace-plan` (plan iterations) · `/pace-retro` (review metrics) · `/pace-biz` (business planning) · `/pace-release` (ship it)\n\n## Core Capabilities\n\n### Change Management (Core Differentiator)\n\n| Capability | Description |\n|-----------|-------------|\n| Requirement changes | Add features, pause, change scope — auto impact analysis, orderly adjustment, Claude won't change the plan unilaterally |\n| Complexity awareness | Auto-assesses task complexity, small changes fast-track, large changes full process, complexity drift auto-detected |\n| Technical debt management | `tech-debt` CR type + iteration capacity reservation + trend tracking |\n| Architecture decisions | `/pace-trace arch` manages cross-CR Architecture Decision Records (ADR) |\n\n### Quality \u0026 Traceability\n\n| Capability | Description |\n|-----------|-------------|\n| Quality gates | Code quality + requirement consistency auto-check + adversarial review, human approval cannot be skipped |\n| Goal traceability | From business goals to code changes, always traceable |\n| Test verification | Requirements-driven — strategy generation, coverage analysis, AI acceptance verification, change impact regression |\n| Semantic drift detection | Continuous monitoring of code-requirement alignment during development, Review includes semantic consistency score |\n\n### Development Rhythm\n\n| Capability | Description |\n|-----------|-------------|\n| Cross-session restore | Session interrupted? Auto-resume, zero re-explanation, experience persists across sessions |\n| Iteration management | Plan → Execute → Review full cycle, auto-recommends next step |\n| Progressive autonomy | Assisted / Standard / Autonomous — more guidance for new users, less for experienced ones |\n| DORA proxy metrics | Deploy frequency / Lead time / Failure rate / MTTR proxy values, Elite~Low benchmarks + trend comparison |\n| CI/CD awareness | Auto-detects CI tool type, Gate 4 auto-queries CI status, zero config |\n| Risk fabric | OWASP-aware security scanning + Pre-flight 5-dimension risk scan + Runtime monitoring + Graduated autonomous response (High requires human confirmation) |\n| Delivery prediction | AI-powered iteration delivery probability forecasting, bottleneck identification, and risk early warning |\n| Cross-project insights | High-confidence insights exportable/importable to other projects, reducing redundant learning |\n\n## Workflow\n\n### Two Modes\n\n- **Explore mode** (default): Freely read code, analyze problems, discuss approaches. No process triggered.\n- **Advance mode** (when changing code): Auto-creates tasks, tracks progress, checks quality. Small changes fast-track, large changes full process.\n\nWhen unsure, Claude asks: \"Ready to start coding, or just exploring?\"\n\n### Workflow\n\n```\nNormal flow:\nStart ──→ In Progress ──→ Pending Review ──→ Done\n              │                │\n        Auto quality check  You approve      Auto merge + status update\n        (Claude handles)   (you decide)\n\nAnytime:\n  Requirements changed ──→ Impact analysis ──→ Adjust plan ──→ Continue\n  Session interrupted  ──→ Next session auto-resumes from where you left off\n\nFull cycle (optional):\n  Plan (pace-plan) → Build (pace-dev) → Review (pace-retro) → Next cycle\n```\n\n## Design Principles\n\n| Principle | Meaning |\n|-----------|---------|\n| Zero friction | Natural language works, no jargon to learn |\n| Progressive disclosure | Default 1-line output, details on demand |\n| Byproducts not prerequisites | Structured data is auto-produced from work, not a required input |\n| Interruption tolerance | Interrupt at any point, seamless resume next time |\n\n## vs Alternatives\n\n| Dimension | GitHub Issues / Manual | devpace |\n|-----------|----------------------|---------|\n| Core model | Task list | Goal → Feature → Code Change traceability |\n| Requirement changes | Manual impact assessment | Auto impact analysis + orderly adjustment |\n| Claude's role | Executor (you direct each step) | Autonomous collaborator (auto-advances, self-checks, waits for your decisions) |\n| Traceability | Task → Code | Business Goal → Feature → Change → Code |\n| Metrics | Completion count | Quality pass rate + value alignment + DORA proxies |\n\n## What devpace is NOT\n\n- **Not a CI/CD pipeline** — it works alongside your existing tools (GitHub Actions, Jenkins, etc.)\n- **Not a project management platform** — no web dashboard, no team features, pure CLI\n- **Not a replacement for git** — it creates Markdown state files in `.devpace/`, your code stays in git\n\n## Feedback\n\nTried devpace? [Tell us what you think](https://github.com/arch-team/devpace/issues/3) — one word or a full report, everything helps.\n\n## Learn More\n\n- [User Guide](docs/user-guide.md) — Full command reference, modes, state machine details ([中文](docs/user-guide_zh.md))\n- [Walkthrough](examples/todo-app-walkthrough.md) — Complete example from init to finish ([中文](examples/todo-app-walkthrough_zh.md))\n- [Contributing](CONTRIBUTING.md) — Dev environment, testing, PR process\n- [Changelog](CHANGELOG.md) — Version history ([English summaries on GitHub Releases](https://github.com/arch-team/devpace/releases))\n- [Troubleshooting](https://github.com/arch-team/devpace/issues?q=label%3Abug) — Search known issues or open a new one\n\n---\nMIT\n","funding_links":[],"categories":[],"sub_categories":[],"project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farch-team%2Fdevpace","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Farch-team%2Fdevpace","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Farch-team%2Fdevpace/lists"}