intervals-to-freshbooks
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell scripts and sqlite3 commands to facilitate data transfer and local logging. These operations are performed using standard system tools and authorized API scripts.
- [EXTERNAL_DOWNLOADS]: The skill interacts with legitimate FreshBooks API and authentication endpoints (api.freshbooks.com and auth.freshbooks.com) as part of its core synchronization functionality.
- [PROMPT_INJECTION]: The skill is subject to indirect prompt injection risks as it processes third-party website content.
- Ingestion points: Data is retrieved from bhi.intervalsonline.com via browser automation.
- Boundary markers: None identified in the interpolation of project names into SQL or shell templates.
- Capability inventory: Access to browser scripting, shell execution, and local database file modification.
- Sanitization: API requests use jq for safety, while SQLite logging relies on the agent manually populating a heredoc template.
Audit Metadata