The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to process untrusted data from a variety of external sources, creating a vulnerability surface for indirect prompt injection attacks where malicious instructions could be embedded in the data being reviewed.
Ingestion points: The workflow requires the agent to perform extensive searches across Google, LinkedIn, news terminals, and corporate registries to compare current findings with original onboarding data (Step 2 in SKILL.md).
Boundary markers: The skill does not define technical boundary markers or specific "ignore embedded instructions" delimiters for the external content; instead, it relies on procedural "HITL Checkpoints" and manual "Analyst Actions" for verification.
Capability inventory: The skill utilizes browser, excel, and pdf connectors to interact with the web and generate reports based on the retrieved information.
Sanitization: There are no explicit instructions for sanitizing or escaping the data fetched from the external searches before it is incorporated into the agent's context or final reports.

refresh