agentleFS
tamldtamld/defense-in-depth/.agents/skills/skill-creator/SKILL.md

Creator

Skill4 starsChanged 7 months agoMITView on GitHub

Multi-branch creation skill for scaffolding new skills, rules, and workflows/procedures. Ensures all created artifacts comply with project standards from day one.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

11 rules. 0 are common, 9 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Create a skill without a flowchart

    line 94
  2. Distinctive:

    Leave placeholder content ("TODO: fill in steps")

    line 96
  3. Distinctive:

    Create a skill that duplicates an existing one

    line 97
  4. Distinctive:

    Every rule MUST map to a cognitive branch

    line 137
  5. Very specific:

    Update the Rules Registry table in COGNITIVE_TREE.md

    line 138
  6. Distinctive:

    Create a rule without mapping it to a cognitive branch

    line 149
  7. Very specific:

    Write advisory-only rules for critical constraints (use enforcement: deterministic)

    line 150
  8. Distinctive:

    Create rules that contradict existing rules without deprecating the old one

    line 151
  9. Distinctive:

    Write vague rules ("be careful with X") — always include concrete examples

    line 152
  10. Distinctive:

    Write steps that say "do the right thing" without specifying what that is

    line 196
  11. Distinctive:

    Create procedures longer than 200 lines (split into sub-procedures if needed)

    line 197

“Common” is measured: kept by at least 6 in 10 of the 5 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

🏗️ Creator — Scaffold Skills, Rules & Workflows

"Every governance artifact starts from a template, but becomes unique through substance."

This skill has 3 branches. Each branch produces a different type of governance artifact, enforcing project standards at creation time so violations are impossible.


Branch Router

flowchart TD
    START["Need to create\na governance artifact"] --> Q1{"What type?"}
    Q1 -->|"Reusable agent\ncapability"| SKILL["Branch A\nSkill Creator"]
    Q1 -->|"Immutable\nproject standard"| RULE["Branch B\nRule Creator"]
    Q1 -->|"Step-by-step\noperational guide"| WORKFLOW["Branch C\nWorkflow Creator"]
    
    SKILL --> VALIDATE["Validate against\nrule-consistency"]
    RULE --> VALIDATE
    WORKFLOW --> VALIDATE
    VALIDATE --> DONE["✅ Ready for review"]

Branch A: Skill Creator

When to Use

  • Agent discovers a repeatable capability that should be formalized
  • User requests a new agent skill
  • Importing a skill from an external source (e.g., skills.sh, GitHub)

Creation Protocol

  1. Name the skill — skill-{name}/SKILL.md
    .agents/skills/skill-{name}/SKILL.md
  2. Generate frontmatter (mandatory fields)
    ---
    name: {Human-readable name}
    description: >
      {What this skill does, when to use it}
    version: 1.0.0
    triggers:
      - {when should this skill activate}
    ---
  3. Write the body (mandatory sections)

    SectionRequiredPurpose
    Overview flowchart✅Visual decision path (mermaid)
    When to Use✅Trigger conditions
    Steps✅Concrete execution steps
    Do Not Do✅Explicit anti-patterns
    Quick Reference⭐Optional cheat sheet
  4. Validate
    • Has mermaid flowchart (rule-consistency §5)
    • No TODO/TBD/PLACEHOLDER (rule-zero-theater)
    • All content in English (rule-adaptive-language)
    • Description is specific, not generic (rule-lesson-quality spirit)

Importing External Skills

flowchart TD
    SOURCE["External skill\n(URL, repo, skills.sh)"] --> FETCH["Fetch/download\nskill content"]
    FETCH --> ADAPT["Adapt to project\nfrontmatter format"]
    ADAPT --> VALIDATE["Validate against\nproject rules"]
    VALIDATE -->|"Pass"| INSTALL["Install to\n.agents/skills/"]
    VALIDATE -->|"Fail"| FIX["Fix violations\nbefore installing"]

Do Not Do

  • Create a skill without a flowchart
  • Copy a skill verbatim from another project without adapting frontmatter
  • Leave placeholder content ("TODO: fill in steps")
  • Create a skill that duplicates an existing one

Branch B: Rule Creator

When to Use

  • A recurring pattern violation needs codification
  • A lesson keeps appearing (> 3 occurrences)
  • A new governance constraint is needed

Creation Protocol

  1. Name the rule — rule-{name}.md
    .agents/rules/rule-{name}.md
  2. Generate frontmatter (mandatory fields)
    ---
    id: RULE-{UPPER-CASE-NAME}
    status: active
    version: 1.0.0
    enforcement: deterministic | advisory
    cognitive_branch: evidence | mechanism | growth | hitl_trunk
    ---
  3. Write the body (mandatory sections)

    SectionRequiredPurpose
    Tagline quote✅One-line philosophical summary
    Decision flowchart✅Visual decision path (mermaid)
    Do / Do Not Do tables✅Explicit behaviors
    Anti-Patterns table✅❌/✅ comparison
    Executable Logic✅Regex for automated detection
  4. Map to COGNITIVE_TREE
    • Every rule MUST map to a cognitive branch
    • Update the Rules Registry table in COGNITIVE_TREE.md
  5. Validate
    • Has YAML frontmatter with all required fields
    • Has mermaid flowchart
    • Has Do-Not-Do section
    • Has executable logic regex
    • Mapped to cognitive branch
    • Listed in COGNITIVE_TREE Rules Registry

Do Not Do

  • Create a rule without mapping it to a cognitive branch
  • Write advisory-only rules for critical constraints (use enforcement: deterministic)
  • Create rules that contradict existing rules without deprecating the old one
  • Write vague rules ("be careful with X") — always include concrete examples

Branch C: Workflow/Procedure Creator

When to Use

  • A multi-step process needs to be documented
  • A repeatable operational task needs standardization
  • An agent keeps asking "how do I do X?"

Creation Protocol

  1. Name the procedure — procedure-{name}.md
    .agents/workflows/procedure-{name}.md
  2. Generate frontmatter
    ---
    description: {Short title of what this procedure does}
    ---
  3. Write the body (mandatory sections)

    SectionRequiredPurpose
    Overview flowchart✅Visual process path (mermaid)
    Prerequisites✅What must be true before starting
    Steps (numbered)✅Concrete, actionable steps
    Verification✅How to confirm success
    Do Not Do✅Common mistakes
    Troubleshooting⭐Optional error recovery
  4. Validate
    • Has mermaid flowchart
    • Steps are numbered and actionable
    • Each step has a verification check
    • No TODO/TBD placeholders

Do Not Do

  • Mix procedure steps with rule definitions
  • Write steps that say "do the right thing" without specifying what that is
  • Create procedures longer than 200 lines (split into sub-procedures if needed)
  • Embed rule content that belongs in .agents/rules/

Download now (ZIP)

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.