registry.update

Purpose

The registry.update skill centralizes all changes to /registry/skills.json. Instead of each skill writing to the registry directly, they call this skill to ensure consistency, policy enforcement, and audit logging.

Usage

Basic Usage

python skills/registry.update/registry_update.py <path_to_skill.yaml>

Arguments

Argument	Type	Required	Description
manifest_path	string	Yes	Path to the skill manifest file (skill.yaml)

Behavior

1. Policy Enforcement: Runs policy.enforce skill (if available) to validate the manifest against organizational policies
2. Load Manifest: Reads and parses the skill manifest YAML
3. Update Registry: Adds or updates the skill entry in /registry/skills.json
4. Thread-Safe: Uses file locking to ensure safe concurrent updates
5. Audit Trail: Records all registry modifications

Outputs

Success Response

{
  "ok": true,
  "status": "success",
  "errors": [],
  "path": "skills/api.validate/skill.yaml",
  "details": {
    "skill_name": "api.validate",
    "version": "0.1.0",
    "action": "updated",
    "registry_file": "/registry/skills.json",
    "policy_enforced": true
  }
}

Failure Response (Policy Violation)

{
  "ok": false,
  "status": "failed",
  "errors": [
    "Policy violations detected:",
    "  - Skill name must follow domain.action pattern",
    "  - Description must be at least 20 characters"
  ],
  "path": "skills/bad-skill/skill.yaml"
}

Policy Enforcement

Before updating the registry, this skill runs policy.enforce (if available) to validate:

• Naming Conventions: Skills follow domain.action pattern
• Required Fields: All mandatory fields present and valid
• Dependencies: Referenced dependencies exist in registry
• Version Conflicts: No version conflicts with existing skills

If policy enforcement fails, the registry update is blocked and errors are returned.

Thread Safety

The skill uses file locking via safe_update_json to ensure:

• Multiple concurrent updates don't corrupt the registry
• Atomic read-modify-write operations
• Proper error handling and rollback on failure

Integration

With skill.define

skill.define automatically calls registry.update after validation:

# This validates AND updates registry
python skills/skill.define/skill_define.py skills/my.skill/skill.yaml

With skill.create

skill.create scaffolds a skill and registers it:

python skills/skill.create/skill_create.py my.skill "Does something"
# Internally calls skill.define which calls registry.update

Direct Usage

For manual registry updates:

python skills/registry.update/registry_update.py skills/custom.skill/skill.yaml

Registry Structure

The /registry/skills.json file has this structure:

{
  "registry_version": "1.0.0",
  "generated_at": "2025-10-23T12:00:00Z",
  "skills": [
    {
      "name": "api.validate",
      "version": "0.1.0",
      "description": "Validate OpenAPI specifications",
      "inputs": ["spec_path", "guideline_set"],
      "outputs": ["validation_report", "valid"],
      "dependencies": ["context.schema"],
      "status": "active",
      "entrypoints": [...],
      "tags": ["api", "validation"]
    }
  ]
}

Files Modified

• Registry: /registry/skills.json – Updated with skill entry
• Logs: Registry updates logged to Betty's logging system

Exit Codes

• 0: Success (registry updated)
• 1: Failure (policy violation or update failed)

Common Errors

Error	Cause	Solution
"Manifest file not found"	Path incorrect or file doesn't exist	Check the path to skill.yaml
"Policy violations detected"	Skill doesn't meet requirements	Fix policy violations listed in errors
"Invalid YAML in manifest"	Malformed YAML syntax	Fix YAML syntax errors

Best Practices

1. Use via skill.define: Don't call directly unless needed
2. Policy Compliance: Ensure skills pass policy checks before registration
3. Version Control: Keep registry changes in git for full history
4. Atomic Updates: The skill handles thread safety automatically

See Also

• skill.define – Validates manifests before calling registry.update (skill.define SKILL.md)
• policy.enforce – Enforces organizational policies (if configured)
• Betty Architecture – Five-Layer Model

Status

Active – Production-ready, core infrastructure skill

Version History

• 0.1.0 (Oct 2025) – Initial implementation with policy enforcement and thread-safe updates