The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands using git, make, npx, and gh (GitHub CLI). These are used for status checking, linting, repository configuration, and deployment operations.
[EXTERNAL_DOWNLOADS]: The skill utilizes npx to execute the skills utility and uses GitHub CLI extensions (gh skill), which involves fetching and running code from the npm registry and GitHub.
[DYNAMIC_EXECUTION]: Employs python3 -c to execute a small Python script for the purpose of validating the syntax of the skills.sh.json configuration file.
[INDIRECT_PROMPT_INJECTION]: The skill processes local configuration data and directory structures to determine repository metadata, creating a potential attack surface for indirect prompt injection.
Ingestion points: Reads data from skills.sh.json, the skills/ directory, and skill file frontmatter.
Boundary markers: Absent for the local configuration data being processed.
Capability inventory: Includes repository visibility modification (gh repo edit), API interactions (gh api), and package installation (npx skills add).
Sanitization: No specific sanitization or validation of the ingested content (beyond JSON syntax checking) is described.

publish-repo