opencli-adapter-author
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is a developer tool focused on reverse-engineering website structures and APIs to create data adapters. It utilizes project-specific tools and operates within a local configuration directory (~/.opencli/).
- [DATA_EXPOSURE]: The skill instructions provide clear guidelines for data de-sensitization, requiring the agent to strip auth-related headers and personal identifiers from raw response samples before saving them as fixtures. This demonstrates a proactive security posture regarding data handling.
- [INDIRECT_PROMPT_INJECTION]: The skill involves ingesting data from third-party websites to extract schema information. While this is an ingestion surface for untrusted data, it is a core functional requirement of the skill, and the workflow emphasizes structured extraction (JSON, CSS selectors) rather than direct instruction execution.
- [COMMAND_EXECUTION]: The skill uses platform-specific Bash tools (
opencli:*) to automate browser-based reconnaissance and verification. These operations are scoped to the development task and utilize approved platform capabilities.
Audit Metadata