exa-webhooks-events
Pass
Audited by Gen Agent Trust Hub on Sep 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data retrieved via the Exa Search API and transmits it to webhook endpoints.
- Ingestion points: External content such as page snippets, titles, and highlights are ingested via
exa.searchAndContentsandexa.findSimilarAndContentsinSKILL.md(Steps 2, 3, and 5). - Boundary markers: No explicit delimiters or instructions are provided to the agent to treat the search results as untrusted data within the provided logic.
- Capability inventory: The skill utilizes the
fetchAPI (network write) to transmit the processed content to awebhookUrland uses standard file system access via allowed-tools. - Sanitization: The implementation uses basic string truncation for snippets but lacks comprehensive sanitization or escaping of the fetched content before transmission.
Audit Metadata