2btwist/dotagents/plan

Claude Code and agent skill by 2btwist/dotagents

Trust tier 3
  • community
Source check pending

Checking that the SKILL.md file is still in the repository.

Create a decision-complete, phased implementation plan after grounding in the code, then write it to specs/plans/YYYY-MM-DD-<slug>.md. Opt-in only: the default route for settled work is a grill, whose decision record `implement` executes directly. Reach for this when the user explicitly asks to plan.

Install

npx vskill@latest install 2btwist/dotagents/plan

SKILL.md

The same SKILL.md was published earlier as 2btwist/dotagents/plan.

Plan

Create a detailed, phased implementation plan by grounding in the code, resolving every open decision, then writing it down. Be skeptical — verify against the repo, don't assume.

Execution style

Execute steps immediately. Don't announce ("I'll now read…"); start with the first tool call. Pause only where a step says to (Step 3 structure confirmation, Step 5 review).

If invoked with a file / ticket / research path, read it FULLY first, then begin. If invoked with nothing, ask for the task (or a path to one) and wait.

Steps

1. Ground in the code

Read every referenced file FULLY (no limit/offset) before anything else. Then spawn the read-only research agents in parallel — codebase-locator (find related files), codebase-analyzer (how it works today), codebase-pattern-finder (patterns to model after). Read everything they surface, fully. Cross-reference the task against what the code actually does; note discrepancies and the true scope.

Done when you can state the current behavior with file:line references, not guesses.

2. Resolve every open decision

Resolve anything the code can't answer with the user. For a substantial plan, interview them down the decision tree: work one branch at a time, resolving each decision before the ones that depend on it, and for every question surface its edge cases and name the tradeoff it makes against the alternative you're rejecting. If a grilling skill is installed, use it for this. For a small plan, just ask the few questions the code couldn't answer.

Then enumerate the installed skills and match each phase's task class to one (ls ~/.claude/skills ~/.codex/skills ~/.pi/agent/skills .claude/skills .agents/skills 2>/dev/null). A locally-installed skill is pre-vetted for this stack — name it in the plan's References and in the relevant phase so /implement reuses it instead of reinventing the approach.

Done when no open questions remain. Every decision is made before you write.

3. Confirm the phase structure

Present the overview and the phase list (each phase's name + what it accomplishes). Get the user's sign-off on phasing and granularity before writing any detail.

4. Write the plan

Write to specs/plans/YYYY-MM-DD-<kebab-slug>.md (create specs/plans/ if needed), following the structure in `PLAN-TEMPLATE.md`.

Done when every phase carries all three: the file paths it touches, the exact commands for its automated success criteria, and the manual items a human must check. A phase whose success criteria are prose rather than a command is not decision-complete — /implement cannot run it.

5. Review

Point the user at the draft and ask two questions: are the phases scoped right, and is any success criterion still too vague to run. Edit in place for each answer.

Done when the user's last round raised no change. Their silence is not agreement — ask.

Security

  • Security scanPassed, 90 of 100Pattern checks on the skill's files.
  • Model reviewPassed

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/2btwist/dotagents/plan/badge)](https://verified-skill.com/skills/2btwist/dotagents/plan)

Something wrong with this skill or its page? Tell us.