agentleFS
thefrederiksenthefrederiksen/cc-director/.claude/skills/skill-manager/skill.md

skill-manager

Skill96 starsChanged 16 days agoMITView on GitHub

Meta-skill for managing Claude Code skills. Create, review, edit, and list skills in a standardized way. Triggers on "/skill-manager", "create skill", "review skill", "list skills", "edit skill", "skill management".

We split this skill into the rules it gives an agent

14 rules. 0 are common, 11 are distinctive, and 3 are very specific to this project.

  • 0

    Common

    Most of the 30 edited versions we read keep these.

  • 11

    Distinctive

    This author's own choices. Worth a look.

  • 3

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create a new skill with proper structure and template.

    line 25
  2. Very specific:

    Check if skill exists: - if .claude/skills/{name}/skill.md exists, abort with error

    line 30
  3. Distinctive:

    Present template options: - ask user to select a template type

    line 31
  4. Very specific:

    Generate skeleton: - create skill.md using selected template

    line 33
  5. Very specific:

    Check .claude/skills/api-validator/ does not exist

    line 59
  6. Distinctive:

    Ask: "Which template should I use for the api-validator skill?"

    line 60
  7. Distinctive:

    Create directory and skill.md from utility template

    line 62
  8. Distinctive:

    Make changes: - use Edit tool to modify

    line 137
  9. Distinctive:

    Ask: "What would you like to edit?"

    line 161
  10. Distinctive:

    Ask: "What should the new description be?"

    line 163
  11. Distinctive:

    Update using Edit tool

    line 164
  12. Distinctive:

    If the requested skill does not exist:

    line 532
  13. Distinctive:

    If the mode is not recognized:

    line 543
  14. Distinctive:

    If skill already exists during create:

    line 554

“Common” is measured: kept by at least 6 in 10 of the 30 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

Skill Manager

A meta-skill for creating, reviewing, editing, and listing Claude Code skills.

Quick Reference

CommandAction
/skill-manager create <name>Create a new skill with selected template
/skill-manager review <name>Validate a skill for quality/completeness
/skill-manager review allReview all skills in the project
/skill-manager edit <name>Edit an existing skill
/skill-manager listList all skills with descriptions
/skill-manager list --categorizeList skills grouped by category

MODE: CREATE

Create a new skill with proper structure and template.

Workflow

  1. Parse skill name from the command argument
  2. Check if skill exists - if .claude/skills/{name}/skill.md exists, abort with error
  3. Present template options - ask user to select a template type
  4. Create directory - .claude/skills/{name}/
  5. Generate skeleton - create skill.md using selected template
  6. Report success - show path to new skill

Template Selection

Present these options to the user:

TemplateBest ForBased On
workflowMulti-step processes with approval checkpointsdeveloper-agent
utilitySimple lookup or single-action toolsgenerate-url
referenceDocumentation-heavy lookup skillssql-query
automationCommand execution and server managementserver-management

Directory Structure Created

.claude/skills/{skill-name}/
    skill.md   (main skill file)

Example

User: /skill-manager create api-validator

Actions:

  1. Check .claude/skills/api-validator/ does not exist
  2. Ask: "Which template should I use for the api-validator skill?"
  3. User selects "utility"
  4. Create directory and skill.md from utility template
  5. Report: "Created skill at .claude/skills/api-validator/skill.md"

MODE: REVIEW

Validate a skill file for quality, completeness, and adherence to standards.

Workflow

  1. Parse skill name from the command argument (or "all")
  2. Read skill file - .claude/skills/{name}/skill.md
  3. Run validation checks - categorized as BLOCKING, WARNING, INFO
  4. Report results - show pass/fail status with details

For "review all"

  1. Find all skills using Glob: .claude/skills/*/skill.md
  2. Run validation on each skill
  3. Summarize results - count of pass/warn/fail

Validation Categories

BLOCKING Issues (must fix before use):

  • YAML frontmatter exists (starts with ---)
  • name: field is present in frontmatter
  • description: field is present and 20+ characters
  • At least one markdown heading (# or ##)
  • File is named skill.md (case-insensitive)

WARNING Issues (should fix):

  • Missing "Triggers" or trigger keywords in description
  • Missing workflow/steps section
  • No version info at end of file
  • No examples section
  • Description over 300 characters (too long for skill list)

INFO Issues (suggestions):

  • Missing quick reference table
  • No error handling section
  • Less than 50 lines (may be too minimal)

Report Format

## Skill Review: {skill-name}

### Status: PASS / WARN / FAIL

### BLOCKING Issues
- None / [list of issues]

### WARNING Issues
- None / [list of issues]

### INFO Suggestions
- None / [list of suggestions]

### Summary
{skill-name} is ready for use / needs attention.

MODE: EDIT

Edit an existing skill with a guided workflow.

Workflow

  1. Parse skill name from the command argument
  2. Read current skill - .claude/skills/{name}/skill.md
  3. Present current content - show frontmatter and structure overview
  4. Ask what to edit - specific section or free-form changes
  5. Make changes - use Edit tool to modify
  6. Run review - validate the changes
  7. Report result - confirm edits were saved

Edit Options

Present these options to the user:

OptionAction
DescriptionUpdate the YAML description field
TriggersAdd/modify trigger keywords
WorkflowUpdate the workflow steps
ExamplesAdd or modify examples
VersionUpdate version number and changelog
CustomMake specific edits by instruction

Example

User: /skill-manager edit commit

Actions:

  1. Read .claude/skills/commit/skill.md
  2. Show: "Current description: 'Invoke the commit skill to create a commit.'"
  3. Ask: "What would you like to edit?"
  4. User selects "Description"
  5. Ask: "What should the new description be?"
  6. Update using Edit tool
  7. Run validation
  8. Report: "Updated commit skill description. Validation: PASS"

MODE: LIST

List all skills with their descriptions.

Workflow

  1. Find all skills using Glob: .claude/skills/*/skill.md
  2. Read each skill - extract name and description from frontmatter
  3. Present results - formatted table

Basic List Output

## Skills ({count} total)

| Skill | Description |
|-------|-------------|
| commit | Create a git commit following... |
| developer-agent | Implement a CenCon issue end-to-end... |
...

Categorized List (--categorize)

Group skills by inferred category based on keywords:

CategoryKeywords in name/description
Developmentbug, fix, code, warning, commit, dead-code
Testingtest, ui-test, qa, review
GitHubissue, pr, pull-request, gh
Documentationdocs, readme, report
Infrastructureserver, build, release
Othereverything else
## Skills by Category

### Development (5)
| Skill | Description |
...

### Testing (3)
| Skill | Description |
...

TEMPLATES

Template: workflow

For multi-step processes with approval checkpoints.

---
name: {skill-name}
description: {Short description}. Triggers on "{trigger1}", "{trigger2}".
---

# {Skill Title}

{Brief overview of what this skill does}

## Quick Reference

| Action | Description |
|--------|-------------|
| Step 1 | {description} |
| Step 2 | {description} |

## CRITICAL: User Approval Required

**NEVER proceed without explicit user approval.**

## Workflow

### Step 1: {Step Name}

{Description of what this step does}

### Step 2: {Step Name}

{Description of what this step does}

### Step 3: Report Completion

Report to user:

{Completion Title}

  • Result: {summary}
  • Files changed: {count}
  • Status: SUCCESS

## Examples

**User:** {example input}

**Agent:**
1. {action 1}
2. {action 2}
3. {result}

---

**Skill Version:** 1.0
**Last Updated:** {date}

Template: utility

For simple lookup or single-action tools.

---
name: {skill-name}
description: {Short description}. Triggers on "{trigger1}", "{trigger2}".
---

# {Skill Title}

{Brief overview of what this skill does}

## CRITICAL: NO GUESSING

**NEVER generate output that is not documented in this skill.**

If the user asks for something not documented:
1. Check the reference code/documentation
2. If not found, respond: "This is not supported."

## Reference Table

| Type | Value |
|------|-------|
| {item1} | {value1} |
| {item2} | {value2} |

## Examples

### Example 1

{example input and output}


## Reference Code

- **Source file:** `{path/to/source.cs}`
- **Documentation:** `{path/to/docs.md}`

---

**Skill Version:** 1.0
**Last Updated:** {date}

Template: reference

For documentation-heavy lookup skills.

---
name: {skill-name}
description: {Short description}. Triggers on "{trigger1}", "{trigger2}".
---

# {Skill Title}

This skill helps you {action description}.

## Connection/Configuration

### {Config Section 1}
- Setting: `{value}`
- Notes: {notes}

### {Config Section 2}
- Setting: `{value}`
- Notes: {notes}

## Usage

### Basic Syntax

```bash
{command example}

Common Options

OptionDescription
-a{description}
-b{description}

Best Practices

  1. {practice 1}
  2. {practice 2}

Workflow

When the user asks to {action}:

  1. Determine target: {question}
  2. Gather parameters: {what to collect}
  3. Execute: {command}
  4. Present results: {format}

Error Handling

If {tool} fails:

  • Check {item 1}
  • Verify {item 2}

Skill Version: 1.0 Last Updated: {date}


### Template: automation

For command execution and server management.

```markdown
---
name: {skill-name}
description: {Short description}. Triggers on "{trigger1}", "{trigger2}".
---

# {Skill Title}

{Brief overview of what this skill manages}

## Quick Reference

| Action | Command |
|--------|---------|
| {action1} | `{command1}` |
| {action2} | `{command2}` |
| Check Status | `{status_command}` |

---

## Options

| Option | Description |
|--------|-------------|
| `--option1` | {description} |
| `--option2` | {description} |

---

## {Action 1}

```bash
{command}

{Description of what this does}

Bash tool parameters: {timeout or run_in_background settings}


{Action 2}

{command}

{Description of what this does}


Configurations

ItemValue
{config1}{value1}
{config2}{value2}

Workflow Examples

{Scenario 1}

{commands}

{Scenario 2}

{commands}

Troubleshooting

{Issue 1}

  • {Solution}

{Issue 2}

  • {Solution}

Skill Version: 1.0 Last Updated: {date} Script/Source: {path/to/script}


---

## Validation Rules Reference

### Frontmatter Requirements

The YAML frontmatter must contain:

```yaml
---
name: skill-name
description: Clear description with trigger keywords. At least 20 characters.
---

Good Description Examples

Good:

  • "Fix GitHub issues by analyzing the issue... Triggers on 'fix bug', 'bug #'."
  • "Execute SQL Server queries on localhost. Use when querying databases, running SQL."
  • "Start and stop servers for testing and development."

Bad:

  • "A skill for stuff" (too vague, no triggers)
  • "skill" (too short)
  • Very long description over 300 characters (gets truncated in skill lists)

Structure Requirements

A well-structured skill should have:

  1. Title heading (# Skill Name)
  2. Quick reference (table of main actions)
  3. Workflow or usage section (how to use)
  4. Examples (at least one)
  5. Version footer (version number, date)

Error Handling

Skill Not Found

If the requested skill does not exist:

Skill '{name}' not found.

Available skills: {list first 10}

Did you mean one of these?

Invalid Mode

If the mode is not recognized:

Unknown mode. Use one of:
- /skill-manager create <name>
- /skill-manager review <name|all>
- /skill-manager edit <name>
- /skill-manager list [--categorize]

Create Conflict

If skill already exists during create:

Skill '{name}' already exists at:
.claude/skills/{name}/skill.md

Use '/skill-manager edit {name}' to modify it.

Implementation Notes

This skill uses only Claude Code tools - no external scripts:

ToolUsage
GlobFind skill files (.claude/skills/*/skill.md)
ReadRead skill content for review/edit/list
WriteCreate new skill files
EditModify existing skill files

Skills Directory

Skills are located at: .claude/skills/

Each skill is a directory containing at minimum a skill.md file.


Skill Version: 1.0 Last Updated: 2025-02-14

Download now (ZIP)

  • Other versions49 projects use a version of this.

    Word for word · 1

    1. thefrederiksen/devthrottle★ 96100% same

    Edited · 0

    None found.

    Wrote their own version · 48

    1. anthropics/claude-for-legal★ 9.5ksame name
    2. buzhangsan/skill-manager★ 321same name
    3. crisandrews/ClawCode★ 63same name
    4. egebese/skill-manager★ 38same name
    5. GeekFamilyCorp/TricorderKit★ 32same name
    6. Starchild-ai-agent/official-skills★ 29same name
    7. Melvynx/ai-maxxing-skills★ 23same name
    8. Interstellar-code/claud-skills★ 17same name
    9. cycleuser/Skills★ 13same name
    10. kargnas/skills★ 9same name

    38 more not listed

    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.

  • Used alongsideProjects with it also use Skill Evolution Manager, github-to-skills and 6 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.
    • Arthur-devroleSkill

      Skill Evolution Manager

      专门用于在对话结束时,根据用户反馈和对话内容总结优化并迭代现有 Skills 的核心工具。它通过吸取对话中的“精华”(如成功的解决方案、失败的教训、特定的代码规范)来持续演进 Skills 库。

      Also in 5 projects that use this

    • Arthur-devroleSkill

      github-to-skills

      Automated factory for converting GitHub repositories into specialized AI skills. Use this skill when the user provides a GitHub URL and wants to "package", "wrap", or "create a skill" from it. It automatically fetches repository details, latest commit hashes, and generates a standardized skill structure with enhanced metadata suitable for lifecycle management.

      Also in 5 projects that use this

    • SentrySkill

      skill-creator

      Create new agent skills following the Agent Skills specification. Use when asked to "create a skill", "add a new skill", "write a skill", "make a skill", "build a skill", or scaffold a new skill with SKILL.md. Guides through requirements, writing, registration, and verification.

      Also in 11 projects that use this

    • OpenAISkill

      brainstorming

      You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.

      Also in 4 projects that use this

    • Microsoft AzureSkill

      humanizer

      Tone enforcement patterns for external-facing community responses

      Also in 3 projects that use this

    • AnthropicSkill

      pdf

      Comprehensive PDF manipulation toolkit for extracting text and tables, creating new PDFs, merging/splitting documents, and handling forms. When Claude needs to fill in a PDF form or programmatically process, generate, or analyze PDF documents at scale.

      Also in 3 projects that use this

  • In this repository30 other files it gives its agents.

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.