pr-review-lite
Skill15 starsChanged 5 months agoMITView on GitHub
Quick checklist review of a PR for style, obvious bugs, and missing tests. Use proactively for lightweight PR reviews.
In this repository21 other files it gives its agents.
- AGENTS.mdAGENTS.md
- public/llms-full.txtllms.txt
- public/llms.txtllms.txt
- add-componentSkill
- add-diagram-componentSkill
- add-languageSkill
- add-pageSkill
- ai-diff-reviewerSkill
- dailybotSkill
- deepworkplanSkill
- doc-editSkill
- fix-lintSkill
- git-commit-pushSkill
- quick-fixSkill
- refactor-safeSkill
- responsive-lighthouse-auditSkill
- security-checkSkill
- translate-syncSkill
- type-fixSkill
- update-stylesSkill
- write-testsSkill
The file
We split this skill into the rules it gives an agent
9 rules. About 0 are common, 8 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈8
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 22
Does NOT perform deep architectural review
- Distinctive: line 23
Does NOT rewrite code
- Distinctive: line 24
Does NOT make changes to the PR
- Distinctive: line 25
Does NOT review security-sensitive changes in depth
- Distinctive: line 53
Read the diff for each file
- Distinctive: line 120
Read-only; no code changes
- Distinctive: line 125
Escalate to Tier 2 (reviewer agent): if:
- Distinctive: line 139
When reviewing Astro code:
- Very specific: line 141
Components: - Check .astro vs .svelte usage is appropriate
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skill: PR Review Lite
Objective
Perform a quick, checklist-based review of a pull request focusing on style issues, obvious bugs, and code quality. This is a lightweight review, not a deep architectural analysis. Adapted for this repo: Astro, Svelte, TypeScript, Biome, AGENTS.md standards.
Non-Goals
- Does NOT perform deep architectural review
- Does NOT rewrite code
- Does NOT make changes to the PR
- Does NOT review security-sensitive changes in depth
Tier Classification
Tier: 1 - Light/Cheap
Reasoning: Checklist-based review with pattern matching, no complex reasoning required. Read-only operation with no code changes.
Inputs
Required Parameters
$PR_FILES: List of files changed in the PR (or PR URL/number)
Optional Parameters
$FOCUS: Specific areas to focus on (default: all)
Prerequisites
- PR diff is accessible
- PR is not draft/WIP
Steps
Step 1: Read the PR
- Get list of changed files
- Read the diff for each file
- Note the scope of changes
Step 2: Run Checklist
Style & Formatting (this repo):
- Consistent formatting (Biome)
- Follows naming conventions (see AGENTS.md)
- No commented-out code
- Import order follows convention
Code Quality:
- No obvious bugs
- TypeScript types used where practical
- Error handling present in API routes
- No hardcoded values that should be config
Astro/Svelte Specific:
- Svelte components use
client:*directive when interactive - Dark mode supported (Tailwind
dark:classes) - Content Collection schema followed
- Pages use Page wrapper pattern (thin wrappers in
src/pages/, shared*Page.astroinsrc/components/pages/) - Page wrappers do not import
MainLayoutdirectly
Tests:
- (Note: Testing not configured in this repo)
- If tests exist, they are meaningful
Documentation:
- Comments for complex logic
- README/docs updated if needed
Step 3: Compile Results
- List all findings
- Categorize by severity (blocking / suggestion)
- Provide file/line references where possible
Output Format
Success Output
## ✅ PR Review Lite Complete
### Summary
{Brief summary of review}
### Findings
**Blocking:** {count}
- {finding 1}
- {finding 2}
**Suggestions:** {count}
- {suggestion 1}
### Checklist
- Style: ✅ / ⚠️ / ❌
- Code Quality: ✅ / ⚠️ / ❌
- Astro/Svelte: ✅ / ⚠️ / ❌
- Docs: ✅ / ⚠️ / ❌Guardrails
Scope Limits
- Read-only; no code changes
- Focus on checklist items; escalate deep architecture to reviewer agent (Tier 2)
Stop Conditions
Escalate to Tier 2 (reviewer agent) if:
- Security-sensitive code involved
- Architectural concerns
- Need for in-depth analysis
Definition of Done
- Checklist completed
- Findings listed with severity
- Clear recommendation (approve / request changes / escalate)
Astro-Specific Items
When reviewing Astro code:
- Components - Check
.astrovs.svelteusage is appropriate - Content - Frontmatter matches schema in
content.config.ts - Pages - Page wrapper pattern (shared
*Page.astrocomponents handleMainLayoutinternally) - API routes - Error handling and response types
- Hydration -
client:loadvsclient:visibleappropriate
Related
- reviewer - For thorough code review
- quick-fix - For small fixes after review
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
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.

