OpenParable
Skip to content
← Back to skills

Plan From Spec

Productivity
plan-from-spec

Plan From Spec is an Agent Skill for productivity published in the majiayu000/claude-skill-registry repository on GitHub: a SKILL.md instruction file an AI agent can load to do this task.

Generate an executable plan + test plan from a spec/decision log, including impact analysis for brownfield changes. Suitable for TDD implementation and PR review.

View on GitHub ↗
QUICK START

How to use this skill

Bring this guide into your coding agent with a prompt tailored to the tool you use.

  1. Open your project in Codex.
  2. Copy and paste the prompt below.
  3. Review files and risks before approving any action.
Prompt to paste
Inspect the Agent Skill at https://github.com/majiayu000/claude-skill-registry/blob/HEAD/skills/planning/plan-from-spec-forgivesam168-ai-dev-workflow/SKILL.md and all supporting files. Report provenance, permissions, scripts, dependencies, and security risks. Propose an exact installation into .agents/skills/plan-from-spec, but wait for my approval before writing files or running scripts.

Copying this prompt does not install or run the skill. Review third-party files before use.

Plan From Spec

When to Use

After a spec exists (either changes/<...>/03-spec.md or specs/<...>/proposal.md).

Inputs

  • Spec path(s)
  • Decision log path (optional but recommended)
  • Constraints (time, scope, dependencies)
  • Brownfield indicator (is this modifying existing system?)

Outputs

  • changes/<...>/04-plan.md — includes:
    • Implementation steps with verification
    • Test strategy
    • Impact analysis (for brownfield/high-risk)
    • Rollback strategy

Impact Analysis (Brownfield/High-Risk)

For any brownfield change or Medium/High risk change, include:

  • Current behavior: What exists today
  • Impacted modules/flows: What will be affected
  • Compatibility/migration: Breaking changes and migration path
  • Regression mitigations: Critical paths to test
  • Observability additions: Logging, monitoring, alerts
  • Rollback strategy: How to revert if needed

Rules

  • Every step must include how to verify
  • Prefer small, reviewable steps (split PRs if needed)
  • Call out risky areas: auth/authz, secrets, data flow, CI/CD
  • Separate refactor from behavior change when possible

Output Format

Provide file-ready markdown for 04-plan.md:

# Implementation Plan: {Feature Name}

## Risk Assessment
- **Risk Level**: Low / Med / High
- **Brownfield**: Yes / No
- **Affected Systems**: {list}

## Impact Analysis (if brownfield/high-risk)
### Current Behavior
{description}

### Impacted Modules
- {module 1}: {impact}
- {module 2}: {impact}

### Rollback Strategy
{how to revert}

## Implementation Steps

### Step 1: {Title}
- **Files**: {paths}
- **Changes**: {description}
- **Verify**: {how to verify}

### Step 2: {Title}
...

## Test Strategy
- **Unit Tests**: {list}
- **Integration Tests**: {list}
- **Regression Tests**: {critical paths}

## Dependencies
- {dependency 1}
- {dependency 2}