The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill includes a self-update mechanism and repository acquisition tools.
The scripts/check_update.py script fetches version and changelog information from the author's official GitHub repository (AI-GO-APP/aigo-template-transfer-skill).
The scripts/acquire.py script utilizes git clone to pull source code from repositories specified by the user as part of the conversion process.
[COMMAND_EXECUTION]: The skill executes shell commands to perform repository management and updates.
It uses the git CLI to clone source repositories and perform fast-forward pulls for skill updates.
All commands are executed using argument lists with subprocess.run to prevent shell injection vulnerabilities.
[INDIRECT_PROMPT_INJECTION]: The skill processes external source code and configuration files which could theoretically contain malicious instructions.
Ingestion points: Files are ingested via scripts/acquire.py into the work/<slug>/template/ directory.
Boundary markers: The SKILL.md instructions explicitly forbid the AI agent from directly modifying files, establishing a 'AI proposes, script applies' workflow. It uses hash-based content gates to detect and block unauthorized changes.
Capability inventory: The skill can write to the file system, execute git commands, and make API calls to developer.ai-go.app and ai-go.app.
Sanitization: The skill uses strict JSON schema validation for templates and requires human review at six distinct manual gates (S0, S1, S3, S5, S6, S9) before any critical action is taken.

aigo-template-transfer