agentleFS
aws-solutionsaws-solutions/konductor/skills/plan-review/SKILL.md

plan-review

Skill11 starsChanged 50 days agoApache-2.0View on GitHub

Critical review of work plans against 4 criteria (clarity, verifiability, context completeness, big picture) with 1-5 star rating and BLOCKED/OKAY/SHIP IT verdicts. Use when evaluating implementation plans, task breakdowns, or work breakdowns before execution.

43 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use plan-execute, pr-comment and 9 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • longranger2Skill

    plan-execute

    Use when the user says "/plan-execute", "plan execute", "implement plan", or "execute plan" and provides a finalized plan file path to carry out. Claude orchestrates, Codex writes code, Claude reviews, and Codex fixes issues until the quality bar is met.

    Also in 3 projects that use this

  • BenjaminGSkill

    pr-comment

    This skill should be used when reviewing someone else's pull request and you want drafted review comments posted to GitHub — either gate / gate-wf findings (tier-marked) or pr-challenge's reviewer questions (unmarked) — each drafted through humanizer and posted as a standalone inline comment at its file:line, after one confirmation. Triggers on "post these as review comments", "comment the gate findings on the PR", "draft and post my review", "leave review comments", after a gate / gate-wf run or a pr-challenge pass on a PR you're reviewing.

    Also in 3 projects that use this

  • MicrosoftSkill

    spec-review

    Independent, multi-model review of a design or spec **before** any code is written, for the microsoft/winappcli repo. Activate when a contributor asks to "review this spec", "review my design", "review this design doc", "validate this approach", "should we build this", "spec review", "design review", or "feature review". Fans out parallel sub-agents — each doing its OWN research against the real codebase and ecosystem rather than trusting the spec — covering necessity & scope, approach & alternatives, feasibility vs reality, risks & unknowns, DX & user impact, and a different-model-family cross-check. Emits a decision-oriented recommendation (proceed / proceed-with-changes / reconsider) to stdout. This is the PRE-CODE companion to the pr-review skill (which reviews code already written); use spec-review at the design/spec stage, not on an implemented diff. Does NOT write code or edit the spec.

    Also in 3 projects that use this

  • AnthropicSkill

    office-hours

    Set up and run Claude Code office hours, including the calendar event when the champion's own calendar connector allows it. Use when a champion wants a regular time for teammates to bring problems.

    Also in 3 projects that use this

  • AWSSkill

    retro

    Conduct a retrospective on a coding session.

    Also in 5 projects that use this

  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 5 projects that use this

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 10 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 35 edited versions we read keep these.

  • 10

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Use this skill when:

    line 16
  2. Distinctive:

    Do NOT use when:

    line 23
  3. Distinctive:

    No vague words like "appropriate", "similar", "etc."

    line 47
  4. Distinctive:

    All required context is in the plan or referenced

    line 75
  5. Distinctive:

    When a plan references files, VERIFY they exist:

    line 93
  6. Very specific:

    Check if path exists (ls <path>)

    line 95
  7. Very specific:

    Check if referenced content exists (grep -n "pattern" <path>)

    line 96
  8. Distinctive:

    Read the full plan — don't start reviewing until you've read everything

    line 121
  9. Distinctive:

    Verify file references — check that referenced files and patterns exist

    line 123
  10. Distinctive:

    Write verdict using the appropriate format (OKAY or BLOCKED)

    line 126
  11. Distinctive:

    No star rating or verdict provided

    line 175
  12. Distinctive:

    No distinction between blocking and non-blocking concerns

    line 182

“Common” is measured: kept by at least 6 in 10 of the 35 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Plan Review

Overview

Evaluates work plans with a critical eye, catching gaps, ambiguities, and missing context that would block implementation. Produces a 1-5 star rating with actionable feedback.

Usage

Use this skill when:

  • Reviewing an implementation plan before greenlighting execution
  • Evaluating a task breakdown from task-decomposition
  • Assessing a feature plan for completeness
  • Running a quality gate on any work plan

Do NOT use when:

  • Reviewing design quality → use design-evaluation
  • Reviewing code → use backend-review, frontend-review, or code-review
  • Reviewing test strategy → use test-coverage-analysis

Core Principle

You are a REVIEWER, not a DESIGNER. The implementation direction in the plan is NOT NEGOTIABLE. Evaluate whether the plan documents that direction clearly enough to execute — NOT whether the direction itself is correct.

4 Evaluation Criteria

1. CLARITY

Does each task specify WHERE to find implementation details?

Pass: Task references specific files, patterns, or examples Fail: Task says "implement X" without pointing to reference code

Checklist:

  • Each task has specific file paths or patterns to follow
  • References to existing code include line numbers or function names
  • Technical terms are defined or exemplified
  • No vague words like "appropriate", "similar", "etc."

2. VERIFICATION

Are acceptance criteria concrete and measurable?

Pass: "Test passes: your test command" or "Build succeeds: your build command" Fail: "Authentication should work correctly"

Checklist:

  • Each task has explicit pass/fail criteria
  • Commands to verify are specified (not "run tests")
  • Expected outputs are described
  • Edge cases have verification steps

3. CONTEXT COMPLETENESS

Less than 10% guesswork required?

Pass: Plan provides all information needed to implement Fail: Implementer must make assumptions or research

Checklist:

  • Dependencies and constraints are listed
  • Related files that might need changes are identified
  • Potential conflicts or side effects are noted
  • All required context is in the plan or referenced

4. BIG PICTURE

Clear purpose, background, and task flow?

Pass: Reader understands WHY this work matters and HOW tasks connect Fail: Tasks are isolated steps without coherent narrative

Checklist:

  • Purpose/objective is stated upfront
  • Tasks flow logically from one to the next
  • Parallel execution opportunities are identified
  • Critical path is clear

File Verification

When a plan references files, VERIFY they exist:

  1. Check if path exists (ls <path>)
  2. Check if referenced content exists (grep -n "pattern" <path>)
  3. Note any discrepancies in the review

Red Flags:

  • References to files that don't exist
  • References to functions/classes not found in the file
  • Outdated file paths (renamed/moved files)

What to Reject

Reject plans that have ANY of these issues:

IssueExampleWhy It's Blocking
Vague tasks"Implement the feature"No guidance on HOW
No acceptance criteria"Make sure it works"No way to verify completion
Missing file references"Update the config"Which config? Where?
Undefined terms"Use the standard pattern"What standard? Show me.
Hand-wavy scope"Handle edge cases"Which ones? All of them?
No contextTask list with no backgroundWhy are we doing this?
Circular references"See task 3" → "See task 1"Infinite loop

Review Workflow

  1. Read the full plan — don't start reviewing until you've read everything
  2. Score each of the 4 criteria — go through checklists, note issues with task numbers
  3. Verify file references — check that referenced files and patterns exist
  4. Identify blocking issues — distinguish "nice to have" from "must fix"
  5. Calculate overall rating using the scoring formula
  6. Write verdict using the appropriate format (OKAY or BLOCKED)

Rating System

RatingVerdictMeaning
⭐BLOCKEDUnusable — fundamental gaps prevent any progress
⭐⭐BLOCKEDMajor revision needed — multiple critical gaps
⭐⭐⭐OKAYWorkable but risky — several unclear areas
⭐⭐⭐⭐OKAYGood plan with minor issues
⭐⭐⭐⭐⭐SHIP ITClear, verifiable, complete — ready for execution

Scoring

Each criterion scores as Pass (fully met), Partial (some items met, some gaps), or Fail (not met).

  • 4 Pass = ⭐⭐⭐⭐⭐
  • 3 Pass, 1 Partial = ⭐⭐⭐⭐
  • 2-3 Pass = ⭐⭐⭐
  • 1 Pass = ⭐⭐
  • 0 Pass = ⭐

Verdict Format

OKAY Verdicts (⭐⭐⭐+)

## VERDICT: OKAY ⭐⭐⭐⭐
### Summary — [1-2 sentence assessment]
### Strengths — [what the plan does well]
### Concerns — [issues to watch during implementation]
### Suggestions — [optional improvements]

BLOCKED Verdicts (⭐-⭐⭐)

## VERDICT: BLOCKED ⭐⭐
### Summary — [why blocked]
### Blocking Issues
1. [CRITICAL] [issue] — What's wrong: ... What's needed: ...
### Must Fix Before Proceeding — [checklist]

Quality Gate

CRITICAL (must fix):

  • Review skips any of the 4 criteria
  • No star rating or verdict provided
  • Verdict contradicts the criteria scores (e.g., all Pass but rated ⭐⭐)

IMPORTANT (should fix):

  • File references not verified
  • Blocking issues lack specific fix instructions
  • No distinction between blocking and non-blocking concerns

SUGGESTION:

  • Could verify more file references
  • Could suggest parallel execution opportunities

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.