review-plan
Review an implementation plan against the current codebase before execution.
Skill by bogdanbujdea
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/bogdanbujdea-review-plan
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Use cases: testing, security
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/review-plan
What's inside
- Skill name
review-plan- Description
- Review an implementation plan against the current codebase before execution. Use when the user asks to review, audit, stress-test, verify, or reality-check a plan for correctness, scope, assumptions, file paths, consistency with existing codebase patterns, risks, edge cases, or validation gaps.
- License
- MIT
- Instructions
- 588 lines in SKILL.md · view on GitHub
Agent Skills spec issues
- body: 588 lines (spec recommends under 500)
Files (1)
- SKILL.md
SKILL.md (first lines)
# Review Plan Review the provided implementation plan impartially before execution. Your job is not to approve the plan. Your job is not to find issues for the sake of finding issues. Your job is to determine whether the plan is accurate, appropriately scoped, consistent with the codebase when available, and safe to execute. Be pragmatic, evidence-based, and honest. ## Core Behavior When this skill is used: 1. Read the full plan before commenting. 2. Identify the stated goal of the plan. 3. Inspect only relevant codebase files before making claims about files, symbols, APIs, tests, commands, architecture, or conventions. 4. Compare the plan against the actual codebase when the codebase is available. 5. Report confirmed problems, uncertainties, design concerns, and things that appear sound. 6. Do not implement the plan. 7. Do not edit files. 8. Do not rewrite the plan unless the user explicitly asks for a revised version. 9. Produce a concise copy/paste summary that the user can paste back into the original planning conversation. Default mode is review-only. ## If Codebase Is Unavailable If the codebase is unavailable, do not make codebase-specific claims. Review only: - goal clarity - internal consistency - scope fit - sequencing - risk - validation Label the result as a plan-only review. Do not claim that files, symbols, commands, tests, routes, components, database fields, or codebase patterns exist or do not exist. ## Goal Clarity Before reviewing details, check whether the plan has a clear goal. The goal is clear only if the plan explains: - what user-visible or externally verifiable behavior should change - what counts as done - the intended scope of the change If the goal is unclear enough that correctness cannot be judged, stop and ask for clarification. If some parts can still be reviewed safely, provide only those observations and clearly label the result as a limited review. Do not review a plan against an inferred goal unless the user explicitly asks you to make assumptions. ## Review Depth Scale the review to the size and risk of the plan. For small plans, keep the review short and focus only on: - goal clarity - codebase accuracy, if codebase is available - consistency with existing patterns, if codebase is available - scope fit - validation For medium or large plans, use the full review structure. For risky plans, pay extra attention to sequencing, reversibility, data safety, authorization, migrations, production behavior, and validation. ## Relevance Rule Only inspect files and behavior directly related to the plan. Do not perform a broad architecture review. Do not search for unrelated improvements. Do not review code style unless the plan depends on it. Do not recommend refactors unless they are necessary for the stated goal. Do not expand the plan beyond what is needed to achieve the goal safely. ## Review Focus Check the plan for the following. ### 1. Goal Clarity Determine whether the goal is specific enough to review. Flag vague goals such as: - "improve the flow" - "clean this up" - "make it better" - "fix the feature" Recommend a clearer goal when possible. ### 2. Codebase Accuracy When the codebase is available, verify that the plan references real codebase elements. Check: - file names - file paths - components - functions - classes - routes - schemas - commands - tests - configuration files - package scripts
From bogdanbujdea/ai-coding-playbook, MIT.
More from bogdanbujdea/ai-coding-playbook
- 0.474score
write-docs
bogdanbujdea
Project documentation toolkit.
SkillCFreshness: CClaude Code +9 - 0.471score
summarize
bogdanbujdea
Create a concise handoff document that summarizes the current conversation for a fresh agent, including goal, context, decisions, references, and next steps.
SkillCFreshness: CClaude Code +9 - 0.471score
summarize-debugging
bogdanbujdea
Create a debugging reset handoff for a fresh agent, preserving the issue, reproduction, impacted files, commands run, attempted fixes, failures, findings, and…
SkillCFreshness: CClaude Code +9 - 0.471score
summarize-for-review
bogdanbujdea
Create a review handoff for completed or partially completed implementation work, summarizing the request, approach, changed files, validation, risks, and…
SkillCFreshness: CClaude Code +9 - 0.450score
to-issues
bogdanbujdea
Break a plan, spec, or PRD into independently-testable issues and publish them to the repo's issue tracker (Azure DevOps, GitHub, or local markdown).
SkillCFreshness: CClaude Code +9 - 0.450score
to-prd
bogdanbujdea
Turn the current conversation context into a PRD saved as a local markdown file.
SkillCFreshness: CClaude Code +9 - 0.445score
grill-me
bogdanbujdea
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree.
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jun 9, 2026 (4 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bogdanbujdea
- 0.474score
write-docs
bogdanbujdea
Project documentation toolkit.
SkillCFreshness: CClaude Code +9 - 0.471score
summarize
bogdanbujdea
Create a concise handoff document that summarizes the current conversation for a fresh agent, including goal, context, decisions, references, and next steps.
SkillCFreshness: CClaude Code +9 - 0.471score
summarize-debugging
bogdanbujdea
Create a debugging reset handoff for a fresh agent, preserving the issue, reproduction, impacted files, commands run, attempted fixes, failures, findings, and…
SkillCFreshness: CClaude Code +9 - 0.471score
summarize-for-review
bogdanbujdea
Create a review handoff for completed or partially completed implementation work, summarizing the request, approach, changed files, validation, risks, and…
SkillCFreshness: CClaude Code +9 - 0.450score
to-issues
bogdanbujdea
Break a plan, spec, or PRD into independently-testable issues and publish them to the repo's issue tracker (Azure DevOps, GitHub, or local markdown).
SkillCFreshness: CClaude Code +9 - 0.450score
to-prd
bogdanbujdea
Turn the current conversation context into a PRD saved as a local markdown file.
SkillCFreshness: CClaude Code +9
Cite this page
review-plan skill. Agent Mart. https://agentmart.dev/skills/bogdanbujdea-review-plan (last updated 2026-10-24)