The Agent Skills Directory

[PROMPT_INJECTION] (SAFE): No instructions attempting to override agent behavior, bypass safety filters, or extract system prompts were detected. The content is strictly limited to technical documentation standards.
[EXTERNAL_DOWNLOADS] (LOW): The skill references standard documentation generators such as typedoc, @microsoft/api-extractor, sphinx, and pdoc. These are well-known tools from trusted package registries (NPM and PyPI). Their mention is instructional and does not involve automated unsafe execution.
[REMOTE_CODE_EXECUTION] (SAFE): There are no patterns involving piping remote content to a shell or executing untrusted code. The bash commands provided are static examples for manual use.
[DATA_EXFILTRATION] (SAFE): No sensitive file paths (e.g., credentials, SSH keys) or network exfiltration patterns were found. Placeholders like jwtSecret and redisUrl are used appropriately within the context of code documentation examples.
[INDIRECT_PROMPT_INJECTION] (INFO): While the skill is designed to have the agent write documentation for code (which could include untrusted content), the skill itself possesses no dangerous capabilities such as filesystem write access, network requests, or subprocess execution that could be exploited through malicious data.

inline-documentation