By alexanderop
Plan-stage-reviewed ticket-to-PR workflow: capture -> research -> plan -> APPROVE -> implement -> validate -> PR. Ports the HumanLayer/CodeLayer methodology to Claude Code. Works with GitHub Issues or Linear, plus parallel research sub-agents and worktree isolation.
Implement an approved plan from thoughts/shared/plans/ phase-by-phase, ticking checkboxes and stopping when reality diverges from the plan. Use after a plan is approved, or via /ticket-flow:implement.
Create a detailed, phased implementation plan with explicit success criteria, saved to thoughts/shared/plans/. Interactive and skeptical. Use after research, or via /ticket-flow:plan.
Document the codebase as it exists today to answer a research question, by spawning parallel read-only sub-agents and synthesizing their findings into thoughts/. Use before planning any non-trivial change, or via /ticket-flow:research.
Run a ticket end-to-end through the plan-stage-reviewed workflow (capture -> research -> plan -> human approval -> implement -> validate -> PR). Works with GitHub Issues or Linear. Use when the user says "ship this ticket", "run the workflow on
Create or refine a tracker ticket (GitHub Issue or Linear) from a rough idea or a thoughts document. Enforces a clear "Problem to solve" before the ticket can be created. Use for "make an issue", "file this ticket", or /ticket-flow:ticket.
External network access
Connects to servers outside your machine
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.
A Claude Code plugin that ports the HumanLayer / CodeLayer methodology: a plan-stage-reviewed ticket-to-PR workflow. Works with GitHub Issues or Linear.
Review and alignment happen at the plan stage, not the PR stage.
📋 Ticket ──/research──▶ 🗺️ Map ──/plan──▶ 📝 Plan ──★ YOU APPROVE ──▶ 💻 Code ──/validate──▶ 🔀 PR
you agent agent (one human gate) agent agent
claude --plugin-dir ./ticket-flow
Then run /reload-plugins after edits. Skills are namespaced under ticket-flow:.
The workflow auto-detects which tracker to use from the ticket reference:
| Reference | Tracker | Status tracked via |
|---|---|---|
#123, owner/repo#123, GitHub URL | GitHub Issues | tf:* labels (one at a time) |
ENG-123, Linear URL | Linear | workflow states |
See references/trackers.md for the exact operations and the
logical-status → label/state mapping. GitHub needs gh authenticated; Linear needs the
linear MCP server (authenticate via /mcp).
/ticket-flow:ship #123 # GitHub issue — whole pipeline, stops for approval after the plan
/ticket-flow:ship ENG-123 # Linear ticket — same pipeline
Or run a single phase:
/ticket-flow:ticket a rough idea # capture -> GitHub issue or Linear ticket (needs a "Problem to solve")
/ticket-flow:research how does auth work # parallel read-only sub-agents -> thoughts/shared/research/
/ticket-flow:plan #123 # phased plan + success criteria -> thoughts/shared/plans/
/ticket-flow:worktree #123 # isolated git worktree (#123 -> issue-123), thoughts/ synced in
/ticket-flow:implement thoughts/shared/plans/foo.md
/ticket-flow:validate thoughts/shared/plans/foo.md
| Plugin piece | CodeLayer equivalent |
|---|---|
skills/* | the .claude/commands/*.md playbooks (research, plan, implement, validate, …) |
skills/ship | the orchestrator that chains them and enforces the human gate |
references/trackers.md | the GitHub-Issues / Linear abstraction (status mapping + commands) |
agents/codebase-researcher.md | the parallel research sub-agents the research command spawns |
.mcp.json (Linear) | the optional Linear MCP integration |
bin/create-worktree.sh | hack/create_worktree.sh — MULTICLAUDE worktree isolation |
hooks/hooks.json | scaffolds the thoughts/ durable-memory directory at session start |
thoughts/shared/ | research + plans persisted outside the context window |
gh CLI, authenticated (gh auth status). Used for issues and PRs.linear MCP server (authenticate via /mcp).git, rsync for worktrees.npx claudepluginhub alexanderop/ticket-flowCreate skills, subagents, hooks, commands, and plugins for Claude Code
Modern Vue 3 development with TypeScript, Composition API, defineModel bindings, Testing Library for user-behavior tests, and MSW for API mocking
Vue language server for Claude Code
A help router plus an optional ship orchestrator for the AFK flow: prototype (throwaway logic/UI exploration) → grill (relentless plan interview, docs-aware) → implement (lead plans, workers run bounded TDD slices) → simplify (parallel cleanup) → qa (frontend/backend evidence verification). Ships a persistent brain/ memory vault (reflect, ruminate, meditate, plan, review, brain) that the flow reads before acting and writes back to as it learns.
Orchestrated multi-agent code review for Claude Code. A /review coordinator fans out to specialized reviewers (security, performance, code quality, docs) via concurrent `claude -p` subprocesses, then dedupes, judges severity, and emits one structured review. Inspired by Cloudflare's orchestrated AI review. Zero runtime dependencies.
Harness-native ECC operator layer - 67 agents, 271 skills, 92 legacy command shims, reusable hooks, rules, selective install profiles, and production-ready workflows for Claude Code, Codex, OpenCode, Cursor, and related agent harnesses
Superpowers Plus core skills library for Claude Code: planning, execution routing, TDD, debugging, and collaboration workflows
Claude harness - A harness for solo developers (Vibecoders) to handle full-cycle contract development.
Core skills library for Claude Code: TDD, debugging, collaboration patterns, and proven techniques
Reliable automation, in-depth debugging, and performance analysis in Chrome using Chrome DevTools and Puppeteer
v9.44.1 — Patch release for Gemini environment/version detection and qwen auth gating. Run /octo:setup.