The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external and potentially untrusted data in the form of code diffs and source files from pull requests. This presents an attack surface where an attacker could embed malicious instructions within code comments or string literals to influence the agent's review logic or tool execution.
Ingestion points: The skill reads pull request diffs and source code files using the Read, Grep, and Glob tools.
Boundary markers: The instructions do not define specific delimiters or "ignore instructions" markers to separate the review guidelines from the content being reviewed.
Capability inventory: The skill has access to powerful tools including Bash, Read, Grep, Glob, and the ability to call other Skill instances, which increases the impact if an injection were to succeed.
Sanitization: There is no evidence of sanitization or filtering applied to the code content before it is processed by the agent.

typescript-review