The Agent Skills Directory

[COMMAND_EXECUTION]: The skill automates publishing workflows by executing system commands such as python3 (for local validation and build scripts), git and gh (for repository and release management), and curl (for triggering cache revalidation webhooks).
[EXTERNAL_DOWNLOADS]: The skill integrates with vendor-specific tools like lov-skill-pricing and uses standard package managers to handle its execution environment. These downloads originate from well-known or vendor-controlled sources.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted input in the form of local skill source files, creating a potential surface for indirect prompt injection.
Ingestion points: Reads SKILL.md, kit.yaml, and other files from user-provided source directories in SKILL.md workflows.
Boundary markers: No specific delimiters or instructions to the model to ignore embedded commands are present when processing source content.
Capability inventory: The skill possesses the ability to write to the file system, perform network requests via curl, and manipulate GitHub repositories via the gh CLI.
Sanitization: The scripts/validate_skill.py script mitigates risks by validating metadata against predefined regex patterns and scanning for absolute user paths to prevent sensitive info leakage.
[CREDENTIALS_UNSAFE]: The skill follows security best practices for secret management by instructing the agent to use environment variables (e.g., LOVSTUDIO_REVALIDATE_SECRET) and existing credential stores for GitHub authentication instead of hardcoding or logging sensitive tokens.

lov-skill-publisher