The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from GitHub issue titles, bodies, and comments. This information is used to make decisions in a triage state machine, creating a vulnerability where a user could embed malicious instructions in an issue.
Ingestion points: SKILL.md describes reading issue bodies and comments during the context gathering phase.
Boundary markers: There are no instructions to use delimiters or ignore embedded instructions within the ingested data.
Capability inventory: The skill has the ability to label issues using the GitHub CLI and run local codebase tests.
Sanitization: No sanitization of external input is defined.
[COMMAND_EXECUTION]: The skill relies on shell command execution for its core functionality.
Evidence: SKILL.md explicitly lists shell commands for gh issue list and mentions running tests to reproduce reported bugs.

process-issues