validate-plugin
Skill0 starsChanged 7 months agoMITView on GitHub
Validate a plugin directory against claude-play marketplace quality standards. Use when reviewing a plugin submission, before committing plugin changes, or when asked to check plugin quality. Checks plugin.json fields, SKILL.md frontmatter, naming conventions, hardcoded paths, and sensitive data patterns.
5 other projects use a version of this.
Adopted by
Most recently taken up by
- radra23/otel-as-code-plugintheir own version, first seen Oct 9, 2026
- grafana/grafana-zabbixtheir own version, first seen Oct 8, 2026
- farzanmrz/meta-devtheir own version, first seen Jul 6, 2026
- langburd/agent-skillstheir own version, first seen Jun 8, 2026
- jpoindexter/theft-kittheir own version, first seen May 20, 2026
Other versions5 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 5
- grafana/grafana-zabbix★ 2.3ksame name
- langburd/agent-skills★ 4same name
- jpoindexter/theft-kit★ 1same name
- radra23/otel-as-code-plugin★ 1same name
- farzanmrz/meta-dev★ 0same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 5 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 4 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Run each check and record the result as OK, WARN, or FAIL.
- Very specific: line 39
Only supported frontmatter attributes: name, description, disable-model-invocation, user-invokable, argument-hint, compatibility, license, …
- Very specific: line 46
Check if plugins/<name>/README.md exists
- Very specific: line 64
Check if a plugin entry with matching name exists in the plugins array
- Very specific: line 85
Use [OK] with 3 spaces, [WARN] with 1 space, [FAIL] with 1 space to align the messages.
- Distinctive: line 89
Do not modify any files — this is a read-only validation skill
- Distinctive: line 90
Do not skip checks — run all 7 even if earlier checks fail
- Distinctive: line 91
Do not fabricate results — if a file can't be read, report the error
- Distinctive: line 92
Prefer dedicated tools (Grep, Glob, Read) over Bash equivalents
“Common” is measured: kept by at least 6 in 10 of the 4 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.
Files
You are a plugin quality validator for the claude-play marketplace.
Arguments
The user may provide a plugin name as an argument (e.g., /validate-plugin my-plugin). If no argument is given, ask which plugin to validate by listing directories under plugins/ (excluding _template).
Workflow
- Resolve plugin path:
plugins/<name>/ - Run all checks in order (see Validation Checks below)
- Print summary using the Output Format below
Validation Checks
Run each check and record the result as OK, WARN, or FAIL.
Check 1: Directory name matches plugin.json name (FAIL)
- Read
plugins/<name>/.claude-plugin/plugin.json - Parse the JSON and compare
namefield to the directory name - FAIL if they don't match or if plugin.json doesn't exist
Check 2: plugin.json has all required fields (FAIL)
- Required fields:
name,description,version,author,license,keywords authormust havenameandemailsubfieldskeywordsmust be a non-empty array- FAIL if any field is missing or empty
Check 3: SKILL.md frontmatter validation (FAIL)
- Glob for
plugins/<name>/skills/*/SKILL.md - Each SKILL.md must have YAML frontmatter (between
---markers) withnameanddescriptionfields - Only supported frontmatter attributes:
name,description,disable-model-invocation,user-invokable,argument-hint,compatibility,license,metadata - WARN if unsupported attributes are found (e.g.,
model,context,allowed-tools) - FAIL if any SKILL.md is missing frontmatter or required fields
- Also FAIL if there are zero SKILL.md files found
Check 4: README.md exists and is non-empty (WARN)
- Check if
plugins/<name>/README.mdexists - WARN if missing or empty (less than 10 characters)
Check 5: No hardcoded absolute paths (WARN)
- Grep all files under
plugins/<name>/for patterns:/home/,/Users/,C:\,/tmp/ - Exclude matches inside code blocks that reference
${CLAUDE_PLUGIN_ROOT} - WARN if any matches found, listing the file and line
Check 6: No sensitive data patterns (FAIL)
- Grep all files under
plugins/<name>/for patterns (case-insensitive):api_key\s*[:=],token\s*[:=],password\s*[:=],secret\s*[:=] - Exclude SKILL.md files that mention these terms in validation check descriptions (like this very file)
- FAIL if any real matches found
Check 7: Plugin registered in marketplace.json (WARN)
- Read
.claude-plugin/marketplace.json - Check if a plugin entry with matching
nameexists in thepluginsarray - WARN if not found (plugin may not be ready for registration yet)
Output Format
Print results in this exact format:
Validating plugin: <name>
[OK] Directory name matches plugin.json name
[FAIL] plugin.json missing required field: keywords
[OK] All SKILL.md files have valid frontmatter
[WARN] README.md is empty
[OK] No hardcoded paths found
[OK] No sensitive data patterns found
[WARN] Plugin not registered in marketplace.json
Result: 4 OK, 2 WARN, 1 FAILUse [OK] with 3 spaces, [WARN] with 1 space, [FAIL] with 1 space to align the messages.
What NOT to Do
- Do not modify any files — this is a read-only validation skill
- Do not skip checks — run all 7 even if earlier checks fail
- Do not fabricate results — if a file can't be read, report the error
- Prefer dedicated tools (Grep, Glob, Read) over Bash equivalents
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.

