Help us improve
Share bugs, ideas, or general feedback.
From deep-dive
Create concrete implementation plan based on research and innovation findings
npx claudepluginhub vm0-ai/team-skills --plugin deep-diveHow this skill is triggered — by the user, by Claude, or both
Slash command
/deep-dive:deep-planThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
You are entering **Deep Plan Mode**. This is the planning phase that transforms research findings and innovation discussions into a concrete implementation plan.
Transforms research findings into actionable implementation plans with granular steps, verification criteria, and stakes-based enforcement. Useful for structuring complex coding tasks before execution.
Generates execution-ready implementation plans from research docs in .artifacts/research/. Ensures junior developers can execute tasks with zero ambiguity. Planning only—no code changes, fixes, or verification.
Creates structured plans for multi-step tasks including software features, implementations, research, or projects. Deepens plans via interactive sub-agent reviews.
Share bugs, ideas, or general feedback.
You are entering Deep Plan Mode. This is the planning phase that transforms research findings and innovation discussions into a concrete implementation plan.
All outputs must be written in English. This includes:
plan.md)This ensures consistency with project standards and accessibility for all contributors.
Before starting:
/tmp/deep-dive/{task-name}/research.md/tmp/deep-dive/{task-name}/innovate.mdIf either document is missing, inform the user and suggest running the appropriate phase first (/deep-dive:deep-research or /deep-dive:deep-innovate).
PERMITTED:
/tmp/deep-dive/{task-name}/plan.mdABSOLUTELY FORBIDDEN:
Apply these thinking approaches during planning:
/tmp/deep-dive/{task-name}/research.md/tmp/deep-dive/{task-name}/innovate.mdBreak down into tasks:
For each task, specify:
Define test requirements:
/testing conventionsIdentify risks and blockers:
/tmp/deep-dive/{task-name}/plan.md$ARGUMENTS
Remember: You are creating a roadmap for implementation. You are NOT implementing yet. The plan should be detailed enough that implementation can proceed smoothly, but no code should be written until the user approves and explicitly requests implementation.