source-command-iterate-plan

Updates existing implementation plans based on user feedback and codebase research.

1|1|Updated Feb 3, 2026
One-click install
npx skills add https://github.com/redblacktree/fastflow --skill source-command-iterate-plan-redblacktree
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: source-command-iterate-plan
Source: https://github.com/redblacktree/fastflow/tree/main/.agents/skills/source-command-iterate-plan
Command: npx skills add https://github.com/redblacktree/fastflow --skill source-command-iterate-plan-redblacktree

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Implementation plans become outdated as requirements change, and manually revising them risks inconsistencies, stale file references, and unverified assumptions. This Skill guides an AI agent through a structured process to update existing plan documents accurately. ## Core Features & Use Cases - Guided Plan Iteration: Parses a plan file path and user feedback, then applies surgical edits while preserving the plan's existing structure and quality standards. - Research-Backed Updates: Spawns parallel research sub-tasks (codebase locators, analyzers, pattern finders) when changes require new technical understanding, ensuring edits are grounded in actual code. - Consistent Success Criteria: Maintains the two-category structure of automated verification (make commands, compilation checks) and manual verification (UI/UX, performance) when updating criteria. - Use Case: A developer has a plan at thoughts/shared/plans/2025-10-16-feature.md and needs to split Phase 2 into backend and frontend phases. The Skill reads the full plan, confirms the intended changes, makes precise edits, and syncs the updated plan. ## Quick Start Ask the agent to iterate on the plan at thoughts/shared/plans/2025-10-16-feature.md by adding a phase for error handling.

Frequently Asked Questions about source-command-iterate-plan

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I update an existing implementation plan with an AI agent?▼

Provide the plan file path (e.g., thoughts/shared/plans/2025-10-16-feature.md) along with your feedback. The agent reads the full plan, optionally researches the codebase, confirms the intended changes with you, then makes precise edits and syncs the result.

How to revise a plan when I only have a plan file path?▼

Invoke the command with just the plan file path and no feedback. The agent confirms it found the plan and prompts you for the specific changes, such as adding a phase, updating success criteria, or adjusting scope.

Does plan iteration require codebase research for every change?▼

No. Research sub-tasks are only spawned when the requested changes require new technical understanding, such as validating code patterns or assumptions. Simple structural edits proceed directly to confirmation and editing.

What happens to success criteria when a plan is updated?▼

The two-category structure is preserved: automated verification uses runnable commands like make test or type checking, while manual verification covers UI/UX and performance. Updates keep criteria measurable and specific.

When should I not use plan iteration?▼

Avoid it when no plan exists yet, since it requires an existing plan file to edit. It is also unsuitable for wholesale rewrites, as the process favors surgical edits that preserve good existing content.