firecrawl-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 is designed to ingest and process untrusted markdown content scraped from external web sources via webhooks.
- Ingestion points: Data enters the system through the
/webhooks/firecrawlendpoint defined inSKILL.md, specifically handlingcrawl.pageandcrawl.completedevent payloads. - Boundary markers: The skill demonstrates HMAC signature verification to authenticate the data source. However, it does not include instructions for sanitizing the scraped content or using delimiters when passing this data to an LLM.
- Capability inventory: The skill code includes logic to store scraped data using
documentStore.upsertandsearchIndex.indexBatch, and the skill's frontmatter allowsWriteandEditoperations. - Sanitization: Verification of the
X-Firecrawl-Signatureheader is correctly implemented to prevent data spoofing, though no content-level filtering is shown for the scraped body.
Audit Metadata