The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external content (specifications, issue bodies, and comments) from potentially untrusted sources via URLs or issue references.
Ingestion points: SKILL.md (Step 1: Gather context) instructions to fetch content from spec paths, issue numbers, or URLs.
Boundary markers: The skill lacks explicit instructions to ignore embedded commands within the fetched content, though it includes specific framing for the resulting ticket generation.
Capability inventory: The skill is capable of writing local files to the .scratch directory and interacting with external issue trackers (GitHub, Linear) if configured via external tools.
Sanitization: No specific sanitization or filtering logic for the ingested content is defined in the instructions.
[REMOTE_CODE_EXECUTION]: No remote code execution patterns, package installations, or piped shell commands were found.
[OBFUSCATION]: No obfuscated content, Base64 strings, or hidden character patterns were detected.

to-tickets