scrapfly-extraction
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted web content (HTML, Markdown, or plain text) through an LLM-based extraction prompt, which is a known vector for indirect prompt injection. An attacker controlling the source content could attempt to manipulate the extracted results.
- Ingestion points: The 'body' parameter in the ExtractionConfig class as shown in SKILL.md.
- Boundary markers: No explicit boundary markers or instructions to ignore embedded commands are included in the provided extraction prompt examples.
- Capability inventory: The skill performs network requests to the Scrapfly API and returns the processed data to the agent context.
- Sanitization: No sanitization or filtering of the ingested content is demonstrated in the implementation examples.
- [EXTERNAL_DOWNLOADS]: The skill documentation instructs the user to install the 'scrapfly-sdk' Python package. This is the official SDK for the Scrapfly service and is a standard requirement for using the tool.
Audit Metadata