The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill points to an external setup guide on GitHub (https://github.com/kixelated/quest/blob/main/SETUP.md) for installing the 'quest' CLI tool. The source repository is not associated with a known trusted organization.
[COMMAND_EXECUTION]: The primary instruction of the skill is to execute the command quest skill complete. The security and behavior of this command are entirely dependent on the externally installed 'quest' binary.
[INDIRECT_PROMPT_INJECTION]: The skill's stated purpose is to decide which open Pull Requests (PRs) to merge. This creates an attack surface where an attacker could include malicious instructions within a PR's title, body, or code to influence the agent's merging logic.
Ingestion points: External Pull Request data retrieved via the quest tool.
Boundary markers: None identified in the skill instructions to delimit untrusted data.
Capability inventory: Execution of shell commands via the quest CLI.
Sanitization: No evidence of sanitization or validation of the PR content before the decision-making process.

quest-complete