The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted external data, creating a surface for indirect prompt injection where instructions hidden in research sources could influence agent behavior.
Ingestion points: External data is ingested via git clone of GitHub repositories and web searches as specified in SKILL.md.
Boundary markers: None identified; the instructions do not include specific delimiters or warnings to ignore instructions within the fetched data.
Capability inventory: Uses git clone, cat, and filesystem commands (mkdir, cd) as documented in SKILL.md to manage and inspect research findings.
Sanitization: None identified; the skill summarizes and presents retrieved evidence to the user without prior sanitization.
[COMMAND_EXECUTION]: The skill uses git clone and cat to retrieve and inspect code. These commands are used for analysis and do not involve execution of untrusted scripts or privilege escalation.
[EXTERNAL_DOWNLOADS]: The skill fetches information from the web and public GitHub repositories to provide technical recommendations and evidence.

research