The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted source code from a repository, which may contain malicious instructions meant to subvert the agent's audit process.
Ingestion points: Local repository files are accessed using Read, Glob, and Grep tools based on the user-supplied [repository-path].
Boundary markers: The skill contains explicit instructions to redact sensitive information such as Authorization, token, callback-token, and PII from diagnostics and logs.
Capability inventory: The skill possesses the ability to modify the analyzed environment using Write and Edit tools.
Sanitization: The instructions mandate constant-time comparisons for webhook signatures and the redaction of sensitive data from all outputs to prevent data leakage during the audit.

salesloft-security-basics