Skip to content
Back to skills

2609 Skill Creation Pr Checklist 59bf371c

ASecurity

**Version:** 1.0.0 **Purpose:** Copy-paste checklist for new skill pull requests **Reference:** [SKILL_SELF_CONTAINMENT_STANDARD.md](SKILL_SELF_CONTAINMENT_STANDARD.md) ---

  • 9 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 11, 2026
developmentjavascripttypescriptpythonrustgojavaphpbashfastapitesting

Works with

  • cli
  • api

Security analysis

A100/100

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 2609-skill_creation_pr_checklist_59bf371c --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 2609 Skill Creation Pr Checklist 59bf371c?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 2609 Skill Creation Pr Checklist 59bf371c
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-2609-skill-creation-pr-checklist-59bf371c/badge)](https://www.skillsdirectory.com/skills/tools-only-2609-skill-creation-pr-checklist-59bf371c)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
# Skill Creation PR Checklist

**Version:** 1.0.0
**Purpose:** Copy-paste checklist for new skill pull requests
**Reference:** [SKILL_SELF_CONTAINMENT_STANDARD.md](SKILL_SELF_CONTAINMENT_STANDARD.md)

---

## Quick Start

Copy this entire checklist into your PR description and check each item before submitting.

---

## New Skill Self-Containment Checklist

**Skill Name:** `[your-skill-name]`
**Category:** `[framework/testing/debugging/collaboration/etc]`
**Toolchain:** `[python/javascript/typescript/rust/go/php/universal/etc]`
**PR Author:** `[@your-github-username]`

---

### ✅ Self-Containment Verification

#### 1. Flat Directory Deployment Test

- [ ] **Copied skill to isolated directory**
  ```bash
  mkdir -p /tmp/skill-test
  cp -r your-skill-name /tmp/skill-test/
  cd /tmp/skill-test/your-skill-name
  ```

- [ ] **Verified all content accessible**
  - SKILL.md displays complete content
  - metadata.json validates
  - No missing files or broken references

- [ ] **No "file not found" errors**
  - All examples work
  - All referenced files exist within skill directory
  - References/ directory (if exists) is accessible

**Isolation Test Result:**
```
✅ PASS - Skill works in flat directory
Location tested: /tmp/skill-test/your-skill-name
All content accessible: YES
Broken references: NONE
```

---

#### 2. Zero Relative Path Violations

- [ ] **Ran grep verification for relative paths**
  ```bash
  cd /Users/masa/Projects/claude-mpm-skills
  grep -r "\.\\./" your-skill-name/
  ```

- [ ] **Output is empty (no violations found)**

- [ ] **Manually verified no path violations**
  - No `../` anywhere in skill
  - No `../../` in any files
  - No hierarchical assumptions

**Grep Verification Output:**
```bash
$ grep -r "\.\\./" your-skill-name/
[paste output here - should be empty]

$ grep -r "from skills\." your-skill-name/
[paste output here - should be empty]

$ grep -r "import.*\.\./" your-skill-name/
[paste output here - should be empty]
```

---

#### 3. Essential Content Inlined

- [ ] **Critical patterns included in SKILL.md**
  - Core functionality patterns (20-50 lines each)
  - Essential code examples
  - No "see other skill" for core features

- [ ] **Setup/configuration complete**
  - Installation instructions
  - Environment setup
  - Configuration examples
  - Dependencies listed

- [ ] **Working examples provided**
  - Examples are complete (not fragments)
  - Examples run independently
  - Examples demonstrate key concepts
  - Comments explain important parts

- [ ] **Common use cases covered**
  - 80% use case addressed
  - Basic patterns included
  - Error handling shown

**Content Completeness Check:**
```
✅ Can user accomplish core task with ONLY this skill? YES
✅ Are critical patterns included (not just referenced)? YES
✅ Do examples work without other skills? YES
✅ Is setup/configuration complete? YES
```

---

#### 4. Complementary Skills Listed Informationally

- [ ] **Used skill names only (no paths)**
  - References use text names: "pytest", "fastapi-local-dev"
  - No file paths: ~~`../../other-skill/SKILL.md`~~
  - No directory paths: ~~`../testing/pytest/`~~

- [ ] **Described relationships clearly**
  - Explained how skills complement each other
  - Noted specific use cases for combinations
  - Clarified integration points (if applicable)

- [ ] **Noted "if deployed" for optional enhancements**
  - Used "if deployed" language
  - Used "optional" or "recommended" language
  - Made clear these are enhancements, not requirements

- [ ] **No "required" or "must install" language**
  - No "This skill requires X"
  - No "Must have Y installed"
  - No "Depends on Z skill"

**Complementary Skills Section Example:**
```markdown
## Complementary Skills

When using this skill, consider (if deployed):

- **skill-name**: How it complements this skill
  - *Use case*: Specific scenario
  - *Integration*: How they work together (optional)

*Note: All skills are independently deployable.*
```

---

#### 5. Graceful Degradation Implemented

- [ ] **Basic functionality is self-contained**
  - Core features work without other skills
  - Essential patterns included inline
  - Minimum viable functionality complete

- [ ] **Advanced features note optional skills**
  - Clearly marked "Advanced" sections
  - Note which skill provides enhancement
  - Explain what additional capability unlocks

- [ ] **Clear distinction: what's included vs. what needs other skills**
  - Inline patterns: "Self-contained pattern (included)"
  - Optional enhancements: "If X skill deployed"
  - Progressive disclosure: Basic → Advanced

**Graceful Degradation Example:**
```markdown
## Testing (Self-Contained)

**Basic testing** (included):
[20-40 lines of inline pattern]

**Advanced fixtures** (if pytest skill deployed):
- Feature 1
- Feature 2

*See pytest skill for comprehensive patterns.*
```

---

#### 6. Tested in Isolation

- [ ] **Verified skill works alone**
  - Deployed to empty directory
  - Read SKILL.md start to finish
  - Confirmed understanding without external references

- [ ] **Examples run without other skills**
  - Copied examples to test environment
  - Ran examples independently
  - Verified no missing imports/dependencies

- [ ] **Documentation makes sense standalone**
  - No unexplained concepts
  - No "see elsewhere" for critical info
  - Complete mental model from SKILL.md alone

- [ ] **No assumptions about other skills being present**
  - Doesn't assume directory structure
  - Doesn't assume other skills deployed
  - Doesn't require bundle deployment

**Isolation Test Evidence:**
```bash
$ cp -r your-skill-name /tmp/skill-test/
$ cd /tmp/skill-test/your-skill-name
$ cat SKILL.md | head -50

[paste first 50 lines - shows content is accessible]

✅ All content loads correctly
✅ No missing references
✅ Examples are complete
✅ Setup instructions present
```

---

#### 7. Bundle Membership Documented

- [ ] **Listed in metadata.json "bundles" field** (if applicable)
  ```json
  {
    "bundles": ["bundle-name"]
  }
  ```

- [ ] **Bundle deployment noted as optional**
  - SKILL.md mentions bundle membership
  - Clarifies bundle is optional
  - Notes skill works outside bundle

- [ ] **Skill works independently of bundle**
  - Tested without other bundle skills
  - No bundle-specific dependencies
  - Standalone functionality complete

**Bundle Membership Section** (if applicable):
```markdown
## Bundle Context

This skill is part of the **[Bundle Name]** bundle.

**If deployed via bundle**, these skills work together:
- skill-1
- skill-2
- skill-3

**If deployed standalone**, this skill is fully functional.

*Bundle deployment is optional.*
```

---

#### 8. Metadata Validation

- [ ] **No other skills in "requires" field**
  ```json
  {
    "requires": []  // Empty or external packages only
  }
  ```

- [ ] **"dependencies" lists only external packages**
  ```json
  {
    "dependencies": ["package-name"]  // External packages, not skills
  }
  ```

- [ ] **"tags" include "self-contained"**
  ```json
  {
    "tags": ["self-contained", "framework", "testing"]
  }
  ```

- [ ] **Version and author specified**
  ```json
  {
    "version": "1.0.0",
    "author": "github-username",
    "updated": "2025-11-30"
  }
  ```

**metadata.json Validation:**
```bash
$ cat your-skill-name/metadata.json | jq .

[paste output here - should show valid JSON]

✅ Valid JSON: YES
✅ No skill dependencies: YES
✅ Self-contained tag: YES
✅ Version specified: YES
```

---

### 🔍 Verification Commands

Run these commands and paste output in PR:

```bash
# Command 1: Check for relative path violations
cd /Users/masa/Projects/claude-mpm-skills
grep -r "\.\\./" your-skill-name/

# Expected output: (empty)

# Command 2: Check for skill import violations
grep -r "from skills\." your-skill-name/

# Expected output: (empty)

# Command 3: Check references/ directory (if exists)
find your-skill-name/references/ -name "*.md" -exec grep -l "\\.\\./\\.\\." {} \;

# Expected output: (empty)

# Command 4: Validate metadata.json
cat your-skill-name/metadata.json | jq .

# Expected output: Valid JSON with required fields

# Command 5: Check for "requires" language
grep -i "requires.*skill\|must.*install\|depends.*on" your-skill-name/SKILL.md

# Expected output: (empty or only "optional"/"if deployed" contexts)
```

---

### 📋 Reviewer Checklist

**For reviewers to verify:**

- [ ] **Grep verification output is empty** (no violations)
- [ ] **SKILL.md is self-sufficient** (read it standalone - makes sense?)
- [ ] **Examples work without other skills** (complete, runnable code)
- [ ] **Complementary skills are informational only** (no paths, no "required")
- [ ] **No relative paths in references/** (if references/ exists)
- [ ] **metadata.json doesn't list skill dependencies** (only external packages)
- [ ] **Content is inlined appropriately** (80% use case covered)
- [ ] **Graceful degradation implemented** (basic vs. advanced clear)

**Common violations to watch for:**
- ❌ `../../other-skill/` paths anywhere
- ❌ "This skill requires X skill" language
- ❌ Missing essential content (just references to other skills)
- ❌ Incomplete examples (fragments, not working code)
- ❌ Hierarchical directory assumptions
- ❌ "See other-skill for setup" without inlining setup

---

### 📝 Additional Context

**Design Decisions:**

[Explain any design choices, content inlining decisions, or trade-offs]

**Content Inlining Rationale:**

[Why you chose to inline certain patterns vs. reference others]

**Complementary Skill Relationships:**

[Explain how this skill relates to others in the ecosystem]

**Testing Approach:**

[Describe how you tested self-containment]

---

### 📚 Reference Documentation

Before submitting, review:

- **[SKILL_SELF_CONTAINMENT_STANDARD.md](SKILL_SELF_CONTAINMENT_STANDARD.md)**: Complete standard
- **[CONTRIBUTING.md](../CONTRIBUTING.md)**: General contribution guidelines
- **[examples/good-self-contained-skill/](../examples/good-self-contained-skill/)**: Template to follow
- **[examples/bad-interdependent-skill/](../examples/bad-interdependent-skill/)**: Anti-patterns to avoid

---

### ✨ Pre-Submission Final Check

Before clicking "Create Pull Request":

- [ ] **All checkboxes above are checked**
- [ ] **All grep commands return empty** (no violations)
- [ ] **Isolation test passed** (works in /tmp/skill-test/)
- [ ] **SKILL.md is self-sufficient** (read it - makes sense alone?)
- [ ] **Examples are complete** (not fragments)
- [ ] **No relative paths anywhere** (verified with grep)
- [ ] **Complementary skills noted informationally** (no paths)
- [ ] **metadata.json validated** (valid JSON, no skill deps)
- [ ] **Graceful degradation implemented** (basic vs. advanced clear)
- [ ] **Bundle membership documented** (if applicable, optional)

---

### 🎯 Success Criteria

This PR is ready to merge when:

✅ All verification commands return expected output (empty or valid)
✅ Skill works in flat directory deployment
✅ No relative path violations found
✅ Essential content inlined (80% use case)
✅ Complementary skills listed informationally only
✅ Graceful degradation implemented
✅ Tested in isolation successfully
✅ Reviewer checklist items verified
✅ CI/CD checks pass

---

## Example: Filled Checklist

**Skill Name:** `fastapi-local-dev`
**Category:** `framework`
**Toolchain:** `python`
**PR Author:** `@example-author`

### ✅ Self-Containment Verification

#### 1. Flat Directory Deployment Test
- [x] Copied skill to isolated directory
- [x] Verified all content accessible
- [x] No "file not found" errors

**Isolation Test Result:**
```
✅ PASS - Skill works in flat directory
Location tested: /tmp/skill-test/fastapi-local-dev
All content accessible: YES
Broken references: NONE
```

#### 2. Zero Relative Path Violations
- [x] Ran grep verification for relative paths
- [x] Output is empty (no violations found)
- [x] Manually verified no path violations

**Grep Verification Output:**
```bash
$ grep -r "\.\\./" fastapi-local-dev/
(empty - no violations)

$ grep -r "from skills\." fastapi-local-dev/
(empty - no violations)
```

#### 3. Essential Content Inlined
- [x] Critical patterns included in SKILL.md
- [x] Setup/configuration complete
- [x] Working examples provided
- [x] Common use cases covered

**Content Completeness Check:**
```
✅ Can user accomplish core task with ONLY this skill? YES
✅ Are critical patterns included (not just referenced)? YES
✅ Do examples work without other skills? YES
✅ Is setup/configuration complete? YES
```

[Continue with remaining checklist items...]

---

## Template Version History

| Version | Date | Changes |
|---------|------|---------|
| 1.0.0 | 2025-11-30 | Initial PR checklist template |

---

**Questions?** Review [SKILL_SELF_CONTAINMENT_STANDARD.md](SKILL_SELF_CONTAINMENT_STANDARD.md) or contact maintainers.

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.