Claude Code Plugins

Community-maintained marketplace

Feedback

Skill Template Example

@AFI-Protocol/afi-skills
0
0

|

Install Skill

1Download skill
2Enable skills in Claude

Open claude.ai/settings/capabilities and find the "Skills" section

3Upload to Claude

Click "Upload skill" and select the downloaded ZIP file

Note: Please verify skill by going through its instructions before using it.

SKILL.md

id skill-template-example
name Skill Template Example
version 0.1.0
domain other
description This is a template for creating new AFI skills. Copy this file to the appropriate domain folder, rename it, and fill in all sections.
inputs [object Object], [object Object]
outputs [object Object], [object Object]
allowed_tools none
risk_level low
determinism_required true
evals [object Object]
owners afi-skills-maintainers
last_updated Fri Nov 21 2025 00:00:00 GMT+0000 (Coordinated Universal Time)
tags template, example, documentation
intent Demonstrate skill template structure and best practices
markets all
timeframes 1h, 4h, 1d
risk_notes - This is a template skill with no actual risk - Real skills should document specific risk considerations
confidence_notes - Template skill has no confidence metrics - Real skills should document accuracy, limitations, and uncertainty sources

Skill Template Example

Purpose

Explain what problem this skill solves and why it exists. Be specific about the use case within the AFI Protocol context.

Example: "This skill validates futures contract specifications to ensure agents correctly interpret margin requirements, settlement times, and funding rate calculations."

When to Use

Describe the triggering conditions or scenarios where an agent should invoke this skill.

Example:

  • When processing a new futures contract listing
  • Before executing a trade that requires margin calculation
  • When validating historical contract data for replay

Workflow

Provide a step-by-step execution flow. Be clear and concise.

  1. Validate inputs: Check that all required parameters are present and well-formed
  2. Load context: Retrieve any necessary reference data (e.g., from TSSD)
  3. Execute core logic: Perform the main computation or analysis
  4. Validate outputs: Ensure results meet expected properties
  5. Return results: Package outputs with metadata

Failure Modes / Edge Cases

Document known failure scenarios and how the skill should handle them.

  • Missing input: Return error with clear message, do not proceed
  • Invalid data format: Attempt parsing with fallback, log warning
  • Determinism violation: If determinism_required: true, fail loudly on any randomness
  • Tool unavailable: Gracefully degrade or return partial results with warning

Example I/O

Provide concrete examples to help agents understand usage.

Example 1: Basic Usage

Input:

{
  "example_input": "sample-value",
  "config": {
    "mode": "strict"
  }
}

Output:

{
  "result": {
    "status": "success",
    "value": 42
  },
  "metadata": {
    "timestamp": "2025-01-15T12:00:00Z",
    "execution_time_ms": 15
  }
}

Example 2: Error Case

Input:

{
  "example_input": null
}

Output:

{
  "result": {
    "status": "error",
    "error": "example_input is required"
  },
  "metadata": {
    "timestamp": "2025-01-15T12:00:00Z"
  }
}

Notes for Validators

(Optional section for skills used in validation workflows)

  • Ensure determinism by seeding any RNG with input-derived values
  • Log all intermediate steps for audit trail
  • Use cryptographic hashes for data integrity checks
  • Report confidence intervals for probabilistic outputs

Implementation Notes

(Optional section for developers)

  • This skill is implemented in afi-core/src/skills/other/skill-template-example.ts
  • Dependencies: None (pure logic)
  • Performance: O(n) where n is input size
  • TODO: Add support for batch processing in v0.2.0