work-with-overleaf
Pass
Audited by Gen Agent Trust Hub on Jul 7, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The script
scripts/overleaf_status.pyexecutes localgitcommands usingsubprocess.runwith argument lists. This is used for inspecting repository state, such as remote URLs and branch names, without exposing the system to shell injection. - [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection surface because it reads and displays project metadata that may originate from untrusted sources.
- Ingestion points: Project file names and Git configuration (remote URLs, branch names) are read by
scripts/overleaf_status.py. - Boundary markers: Data output from the status script is not wrapped in specific delimiters to distinguish it from agent instructions.
- Capability inventory: The skill can execute local Git commands and interact with the file system.
- Sanitization: The status script includes a
redact()function that uses a regular expression to strip authentication tokens from URLs before they are presented to the agent or user.
Audit Metadata