The Agent Skills Directory

[SAFE]: The skill functions as a procedural guide for security auditing, providing checklists for common vulnerabilities such as injection, broken access control, and cryptographic failures.
[DATA_EXPOSURE]: The skill includes regular expression patterns for identifying secrets like AWS keys, GitHub tokens, and Slack tokens. These are provided as analysis tools for the agent to use when reviewing other code, rather than being hardcoded credentials or attempts to leak information.
[INDIRECT_PROMPT_INJECTION]: The skill's primary function is to ingest and analyze untrusted external data (code diffs provided by users). While it lacks specific boundary markers to prevent the analyzed code from influencing the agent, the skill possesses no dangerous capabilities (such as network access, file writing, or command execution) that could be exploited via such an injection.

review-security