skills/abilityai/abilities/trinity-compatibility

trinity-compatibility

SKILL.md

Trinity Compatibility Analysis

Analyze the current agent directory against Trinity Compatible Agent requirements and produce a detailed compatibility report with actionable remediation steps.

Trinity Requirements Reference

Required Files (5 Essential)

  1. template.yaml - Agent metadata with: name, display_name, description, resources.cpu, resources.memory, credentials
  2. CLAUDE.md - Domain-specific instructions (the agent's "brain")
  3. .mcp.json.template - MCP server config using ${VAR} placeholder syntax (if using MCP servers)
  4. .env.example - Documentation of required environment variables (no actual values)
  5. .gitignore - Security-critical exclusions

Required Directory Structure

agent/
├── .git/
├── .gitignore
├── CLAUDE.md
├── README.md
├── template.yaml
├── .claude/
│   ├── agents/
│   ├── commands/
│   ├── skills/
│   ├── skills-library/
│   └── settings.local.json
├── .mcp.json.template
├── .env.example
├── docs/
├── outputs/           # COMMITTED - smaller deliverables
├── content/           # NOT COMMITTED - large generated assets
├── scripts/
└── resources/

.gitignore Must Exclude

Never commit (credentials/secrets):

  • .mcp.json
  • .env
  • *.pem, *.key
  • Credential files

Never commit (generated/large):

  • content/ directory
  • .claude/projects/
  • .claude/statsig/
  • .claude/todos/
  • .claude/debug/

Always commit:

  • .claude/commands/
  • .claude/skills/
  • .claude/agents/
  • settings.local.json
  • outputs/ directory

template.yaml Required Fields

name: lowercase-with-hyphens
display_name: Human Readable Name
description: |
  Multi-line purpose statement
resources:
  cpu: "2"
  memory: "4g"
credentials:
  mcp_servers:
    server_name:
      - VAR_NAME
  env_file:
    - VAR_NAME

Security Requirements

  • No hardcoded credentials in any file
  • All secrets excluded via .gitignore
  • Placeholder syntax ${VAR_NAME} in .mcp.json.template
  • Never push secrets to GitHub

Analysis Task

  1. Read current agent files:

    • Check for existence of all 5 required files
    • Read template.yaml and verify required fields
    • Read .gitignore and check for required exclusions
    • Check .mcp.json for hardcoded credentials vs .mcp.json.template with placeholders
    • Scan for any .env files that shouldn't exist
  2. Verify directory structure:

    • List top-level directories
    • Check for required directories (outputs/, content/, .claude/)
    • Verify .claude/ subdirectories
  3. Security audit:

    • Grep for potential hardcoded credentials (API keys, tokens, passwords)
    • Check if .mcp.json exists (should only be .mcp.json.template)
    • Verify .env.example exists but .env does not (or is gitignored)
  4. Generate compatibility report:

## Trinity Compatibility Report

### Status: [COMPATIBLE / NEEDS WORK]

### Required Files
| File | Status | Notes |
|------|--------|-------|
| template.yaml | [X] / [ ] | ... |
| CLAUDE.md | [X] / [ ] | ... |
| .mcp.json.template | [X] / [ ] | ... |
| .env.example | [X] / [ ] | ... |
| .gitignore | [X] / [ ] | ... |

### Directory Structure
| Directory | Status | Notes |
|-----------|--------|-------|
| .claude/ | [X] / [ ] | ... |
| outputs/ | [X] / [ ] | ... |
| content/ | [X] / [ ] | ... |
| scripts/ | [X] / [ ] | ... |

### Security Check
| Item | Status | Notes |
|------|--------|-------|
| No hardcoded credentials | [X] / [ ] | ... |
| .mcp.json excluded | [X] / [ ] | ... |
| .env excluded | [X] / [ ] | ... |
| Proper .gitignore | [X] / [ ] | ... |

### Required Actions
1. [Action item with specific file and change needed]
2. [Another action item]
...

### Optional Improvements
- [Nice-to-have improvements]
  1. Wait for user approval before making any changes

Present the report and ask: "Would you like me to implement these changes to make this agent Trinity-compatible?"

Weekly Installs
1
First Seen
11 days ago
Installed on
mcpjam1
claude-code1
junie1
windsurf1
zencoder1
crush1