name: autonomous-gap-detector description: Protocol for agents to autonomously detect skill gaps and maintain documentation health.

Autonomous Gap Detector Protocol

This protocol defines how an AI agent should autonomously manage skill gaps within a repository.

🤖 Agent Instructions (CRITICAL)

As an AI Agent, you are responsible for maintaining the "Documentation Health" of the project. You MUST trigger the Gap Detection process in the following scenarios:

1. Task Initialization: When starting a new feature, scan for existing skills to see if you have the necessary knowledge.
2. Dependency Changes: Immediately after adding or updating any library (e.g., in package.json, requirements.txt).
3. Task Completion (DoD Check): Before marking any task as "Done", you must verify that no new skill gaps were introduced.

Execution Protocol

When triggered, you must perform these steps:

1. Run Scanner: Execute the diagnostic script:

   python d:\Cerebra\cerebraSkills\scripts\skill_watcher.py --target .
   

2. Analyze GAP_REPORT.md:

   • If RED GAPS are found:
     • Analyze if the library is critical for the project's long-term maintenance.
     • If it represents a new logic pattern, use the Skill Architect (72-metacognitive-skill-architect) to draft a new skill.
     • Ask the user if they want you to finalize the new skill now or later.
   • If ALL GREEN: Proceed with the task.

3. Cross-Reference: Check if the library is already in LIB_MAPPING in scripts/skill_watcher.py. If not, update the mapping to ensure future scans are accurate.

Reporting Logic

Any gaps identified should be reported to the user in this format:

"🚨 Skill Gap Detected: You are using {library}, but we don't have a standardized Skill for it yet. This may lead to maintenance issues. Should I draft a Skill for this now?"

Technical Details

Scanner Script

• Location: d:\Cerebra\cerebraSkills\scripts\skill_watcher.py
• Config: Edit LIB_MAPPING to add new technology associations.

Logic Flow

1. Parsing project dependency files.
2. Filtering against IGNORE_FILES and IGNORE_DIRS.
3. Matching against SKILL_INDEX.md entries.
4. Outputting markdown report to GAP_REPORT.md.