The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute curl commands to perform API requests to api.github.com. These commands include headers for authorization using the GITHUB_TOKEN environment variable.
[INDIRECT_PROMPT_INJECTION]: The skill processes data returned from the GitHub API, which represents a potential attack surface if organization names, usernames, or other metadata contain malicious instructions.
Ingestion points: JSON responses from GitHub API endpoints (e.g., /copilot/billing/seats, /copilot/metrics) processed in SKILL.md.
Boundary markers: None present; the agent is expected to parse raw JSON responses directly.
Capability inventory: Shell command execution via curl and file writing to /tmp for request bodies.
Sanitization: No explicit validation or filtering is performed on the data retrieved from the external API before it is consumed by the agent.

github-copilot