The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from external websites during research phases. It mitigates potential injection risks by instructing the agent to treat external content as untrusted and ignore embedded commands. Evidence Chain: 1. Ingestion points: External URLs (guidelines, articles) read by browser tools. 2. Boundary markers: Explicit instructions to treat page content as untrusted. 3. Capability inventory: Web browser interaction and email submission tools. 4. Sanitization: Mandatory manual approval gate for all external actions.
[DATA_EXPOSURE_AND_EXFILTRATION]: The skill manages local campaign logs and project context files but includes clear rules against storing credentials, passwords, or sensitive personal data within these records.

backlink-outreach