webflow-observability
Pass
Audited by Gen Agent Trust Hub on Sep 10, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from the local repository using Read, Glob, and Grep tools, which could potentially contain malicious instructions targeting the agent.\n
- Ingestion points: Repository manifests, instructions, and integration paths are read using repository inspection tools as described in SKILL.md.\n
- Boundary markers: The skill explicitly defines 'Approval Boundaries' that mandate fresh user authorization for any creation, update, deletion, or production deployment.\n
- Capability inventory: The agent is permitted to use Edit, Write, and WebFetch tools to modify files and fetch external documentation.\n
- Sanitization: Workflow instructions require redacting site identities and hashing sensitive dimensions, while explicitly forbidding the persistence or logging of authentication tokens.\n- [SAFE]: External references are directed exclusively to well-known official developer documentation and API resources for Webflow.
Audit Metadata