eliasoulkadi/shokunin/skill-creator

Skill Creator

A skill for creating new skills and iteratively improving them.

At a high level, the process of creating a skill goes like this:

• Decide what you want the skill to do and roughly how it should do it
• Write a draft of the skill
• Create a few test prompts and run the skill on them
• Help the user evaluate the results both qualitatively and quantitatively
• Rewrite the skill based on feedback from the user's evaluation of the results
• Repeat until you're satisfied
• Expand the test set and try again at larger scale

Your job when using this skill is to figure out where the user is in this process and then jump in and help them progress through these stages. So for instance, maybe they're like "I want to make a skill for X". You can help narrow down what they mean, write a draft, write the test cases, figure out how they want to evaluate, run all the prompts, and repeat.

On the other hand, maybe they already have a draft of the skill. In this case you can go straight to the eval/iterate part of the loop.

Of course, you should always be flexible and if the user is like "I don't need to run a bunch of evaluations, just vibe with me", you can do that instead.

Then after the skill is done (but again, the order is flexible), you can also optimize the description to improve the triggering of the skill.

Cool? Cool.

Communicating with the user

The skill creator is liable to be used by people across a wide range of familiarity with coding jargon. If you haven't heard (and how could you, it's only very recently that it started), there's a trend now where the power of LLMs is inspiring plumbers to open up their terminals, parents and grandparents to google "how to install npm". On the other hand, the bulk of users are probably fairly computer-literate.

So please pay attention to context cues to understand how to phrase your communication! In the default case, just to give you some idea:

• "evaluation" and "benchmark" are borderline, but OK
• for "JSON" and "assertion" you want to see serious cues from the user that they know what those things are before using them without explaining them

It's OK to briefly explain terms if you're in doubt, and feel free to clarify terms with a short definition if you're unsure if the user will get it.

Workflow

Capture Intent

Start by understanding the user's intent. The current conversation might already contain a workflow the user wants to capture (e.g., they say "turn this into a skill"). If so, extract answers from the conversation history first — the tools used, the sequence of steps, corrections the user made, input/output formats observed. The user may need to fill the gaps, and should confirm before proceeding to the next step.

1. What should this skill enable the model to do?
2. When should this skill trigger? (what user phrases/contexts)
3. What's the expected output format?
4. Should we set up test cases to verify the skill works? Skills with objectively verifiable outputs (file transforms, data extraction, code generation, fixed workflow steps) benefit from test cases. Skills with subjective outputs (writing style, art) often don't need them. Suggest the appropriate default based on the skill type, but let the user decide.

Interview and Research

Proactively ask questions about edge cases, input/output formats, example files, success criteria, and dependencies. Wait to write test prompts until you've got this part ironed out.

Check available tools — if useful for research (searching docs, finding similar skills, looking up best practices), research in parallel via subagents if available, otherwise inline. Come prepared with context to reduce burden on the user.

Write the SKILL.md

Based on the user interview, fill in these components:

• name: Skill identifier
• description: When to trigger, what it does. This is the primary triggering mechanism - include both what the skill does AND specific contexts for when to use it. All "when to use" info goes here, not in the body. Note: currently models have a tendency to "undertrigger" skills — to not use them when they'd be useful. To combat this, make the skill descriptions a little bit "pushy". So for instance, instead of "How to build a simple fast dashboard to display internal data", you might write "How to build a simple fast dashboard to display internal data. Make sure to use this skill whenever the user mentions dashboards, data visualization, internal metrics, or wants to display any kind of company data, even if they don't explicitly ask for a 'dashboard.'"
• compatibility: Required tools, dependencies (optional, rarely needed)
• **the rest of the skill :)

Skill Writing Guide

Anatomy of a Skill

skill-name/
├── SKILL.md (required)
│   ├── YAML frontmatter (name, description required)
│   └── Markdown instructions
└── Bundled Resources (optional)
    ├── scripts/    - Executable code for deterministic/repetitive tasks
    ├── references/ - Docs loaded into context as needed
    └── assets/     - Files used in output (templates, icons, fonts)

Progressive Disclosure

Skills use a three-level loading system:

1. Metadata (name + description) - Always in context (~100 words)
2. SKILL.md body - In context whenever skill triggers (<500 lines ideal)
3. Bundled resources - As needed (unlimited, scripts can execute without loading)

These word counts are approximate and you can feel free to go longer if needed.

Key patterns:

• Keep SKILL.md under 500 lines; if you're approaching this limit, add an additional layer of hierarchy along with clear pointers about where the model using the skill should go next to follow up.
• Reference files clearly from SKILL.md with guidance on when to read them
• For large reference files (>300 lines), include a table of contents

Domain organization: When a skill supports multiple domains/frameworks, organize by variant:

cloud-deploy/
├── SKILL.md (workflow + selection)
└── references/
    ├── aws.md
    ├── gcp.md
    └── azure.md

The model reads only the relevant reference file.

Principle of Lack of Surprise

This goes without saying, but skills must not contain malware, exploit code, or any content that could compromise system security. A skill's contents should not surprise the user in their intent if described. Don't go along with requests to create misleading skills or skills designed to facilitate unauthorized access, data exfiltration, or other malicious activities. Things like a "roleplay as an XYZ" are OK though.

Writing Patterns

Prefer using the imperative form in instructions.

Defining output formats - You can do it like this:

## Report structure
ALWAYS use this exact template:
# [Title]
## Executive summary
## Key findings
## Recommendations

Examples pattern - It's useful to include examples. You can format them like this (but if "Input" and "Output" are in the examples you might want to deviate a little):

## Commit message format
**Example 1:**
Input: Added user authentication with JWT tokens
Output: feat(auth): implement JWT-based authentication

Writing Style

Try to explain to the model why things are important in lieu of heavy-handed musty MUSTs. Use theory of mind and try to make the skill general and not super-narrow to specific examples. Start by writing a draft and then look at it with fresh eyes and improve it.

Test Cases

After writing the skill draft, come up with 2-3 realistic test prompts — the kind of thing a real user would actually say. Share them with the user: "Here are a few test cases I'd like to try. Do these look right, or do you want to add more?" Then run them.

Save test cases to a JSON file in your workspace. Don't write assertions yet — just the prompts. You'll draft assertions in the next step.

{
  "skill_name": "example-skill",
  "evals": [
    {
      "id": 1,
      "prompt": "User's task prompt",
      "expected_output": "Description of expected result",
      "files": []
    }
  ]
}

The full eval schema also supports an assertions array (added later):

{
  "assertions": [
    {
      "id": 1,
      "description": "Output file exists and is valid JSON",
      "type": "file_exists",
      "expected": "output.json"
    }
  ]
}

Running and evaluating test cases

This section is one continuous sequence — don't stop partway through.

Step 1: Run all test cases

For each test case, run the skill against the prompt. If your platform supports subagents/parallel execution, run all test cases simultaneously. Otherwise, run them sequentially.

With-skill run: Execute each test prompt using the skill you're testing. Save all outputs (generated files, logs, console output) to a workspace directory.

Baseline run: Run the same prompts without the skill (or with the old version if you're improving an existing skill). Save outputs to a separate directory.

Organize results by iteration:

workspace/
├── iteration-1/
│   ├── eval-1-with-skill/
│   │   └── outputs/
│   ├── eval-1-baseline/
│   │   └── outputs/
│   └── ...
└── iteration-2/
    └── ...

Create an eval_metadata.json for each test case:

{
  "eval_id": 0,
  "eval_name": "descriptive-name-here",
  "prompt": "The user's task prompt",
  "assertions": []
}

Step 2: Draft assertions

While runs are in progress, draft quantitative assertions for each test case and explain them to the user. If assertions already exist from a previous iteration, review them and explain what they check.

Good assertions are objectively verifiable and have descriptive names — they should read clearly so someone glancing at the results immediately understands what each one checks. Subjective skills (writing style, design quality) are better evaluated qualitatively — don't force assertions onto things that need human judgment.

Update the eval_metadata.json files with the assertions once drafted.

Step 3: Grade and compare results

Once all runs complete:

1. Grade each run — evaluate each assertion against the actual outputs. For assertions that can be checked programmatically, write and run a quick script rather than eyeballing it — scripts are faster, more reliable, and can be reused across iterations.

Save results to grading.json in each run directory: ``json { "expectations": [ { "text": "Output file exists and is valid JSON", "passed": true, "evidence": "File output.json exists and parses as valid JSON with 3 keys" } ] } ``

1. Aggregate into benchmark — compile results into a benchmark.json with pass_rate, time, and tokens for each configuration, showing mean ± stddev and the delta between with-skill and baseline. Place each with_skill version before its baseline counterpart.

1. Analyze patterns — read the benchmark data and surface patterns the aggregate stats might hide. Look for:

• Assertions that always pass regardless of skill (non-discriminating)
• High-variance evals (possibly flaky)
• Time and token tradeoffs (faster but more tokens? worth it?)
• Cases where the skill made things worse

1. Present results to the user — show both qualitative outputs (what files were produced) and quantitative data (benchmark stats). Let the user click through each test case and review outputs. Ask for inline feedback: "How does this look? Anything you'd change?"

Step 4: Read the feedback

When the user tells you they're done reviewing, collect their feedback. Structure it as:

{
  "reviews": [
    {"run_id": "eval-1-with-skill", "feedback": "the chart is missing axis labels"},
    {"run_id": "eval-2-with-skill", "feedback": ""},
    {"run_id": "eval-3-with-skill", "feedback": "perfect, love this"}
  ]
}