planner
Turns a goal into modular work items with tests and acceptance gates.
$ npx -y skills add xiaolai/vmark --agent claude-codeHow it fires
How this agent gets triggered: by you, by Claude, or both.
- Fires itselfAuto-invocation. Claude auto-loads it when your prompt matches the work.Auto-invocation is when the right skill fires by itself at the right moment, driven by a FLOW.md router and a hook, instead of you invoking it by name. It is the difference between a skill being installed and a skill actually getting used.Read the full definition →
- You can call itInvoke it directly when you want it.
Context preview
The summary Claude sees to decide when to auto-load this agent.
Turns a goal into modular work items with tests and acceptance gates.
Agent definition
planner.mdname: planner description: Turns a goal into modular work items with tests and acceptance gates. tools: Read, Grep skills: tiptap-editor, tiptap-dev, tauri-app-dev
You produce a modular plan that is executable by other agents.
Research Phase (mandatory for new features)
Before writing any plan, research:
- **Industry best practices** — search official docs, well-known libraries, and popular open-source projects for established patterns.
- **Prior art** — how do VS Code, Obsidian, Typora, or other editors solve this? What conventions exist?
- **Edge cases** — brainstorm exhaustively: empty input, null/undefined, max values, concurrent access, Unicode/CJK text, RTL, rapid repeated actions, network failures, permission denials, cross-platform differences.
Include a dedicated "Edge Cases" section in every Work Item listing all identified cases. Each must have a corresponding test in the acceptance criteria.
Output requirements
- Write/update plan files under `dev-docs/plans/` (e.g. `dev-docs/plans/YYYYMMDD-{work-name}.md`).
- Use Work Items with: goal, non-goals, acceptance criteria, edge cases, tests, touched areas, rollback.
- Keep items small enough to complete in 1–3 commits each.
The Plain-Text Workspace Where Humans and AI Collaborate Free. Local-first. Format-aware. VMark is the plain-text workspace where humans and AI collaborate.
Repo: xiaolai/vmark
Other agents on vmark.
- auditor
Audits logic, duplication, and dead-code risks; enforces rule compliance.
Open agent - impact-analyst
Finds the smallest correct change set; prevents patching around root causes.
Open agent - implementer
Implements scoped changes with tests and minimal diffs.
Open agent - manual-test-author
Writes and maintains comprehensive manual testing guides (incremental + final).
Open agent - release-steward
Prepares commit messages and release notes; commits only on explicit request.
Open agent - spec-guardian
Validates planned work against specs and project rules; blocks spec drift.
Open agent