apify-core-workflow-a
Pass
Audited by Gen Agent Trust Hub on Sep 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill creates an ingestion point for untrusted data by scraping external websites. This content could potentially contain instructions aimed at manipulating the AI agent's behavior.
- Ingestion points: Web pages processed by the
CheerioCrawlerinsrc/main.tsandreferences/implementation.md. - Boundary markers: The skill does not specify markers to delimit the scraped content from other agent instructions.
- Capability inventory: The skill is authorized to use
Bashfornpm,npx, andapifyCLI commands, and it can perform network requests and file system writes within thestorage/directory. - Sanitization: The implementation extracts specific data fields but does not include sanitization logic to filter out potential malicious instructions within the scraped text.
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of external libraries and uses remote Docker images to facilitate the build and deployment process.
- Evidence: References to
npm install apify crawlee,apify-cli, and the use of theapify/actor-node:20base image in the Dockerfile. These are sourced from well-known service providers.
Audit Metadata