The Agent Skills Directory

[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface because its workflow depends on ingesting and processing untrusted data from external environments.
Ingestion points: Data enters the agent's context through user-provided company lists (Step 0) and results from automated web searches for funding announcements (Step 1).
Boundary markers: The instructions do not define delimiters or specific system instructions to ignore potential commands that could be embedded within the external company data or web search snippets.
Capability inventory: The skill utilizes web-search and has the ability to write campaign packages to the local file system, which could be exploited if malicious instructions are successfully injected via external content.
Sanitization: There is no evidence of sanitization or schema validation mentioned to filter or escape content retrieved from the web before it is used for automated reasoning and email drafting.

funding-signal-outreach