overleaf
Pass
Audited by Gen Agent Trust Hub on Mar 10, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill downloads and installs the
olcliCLI tool from the npm registry and a Homebrew tap during the installation phase. - [COMMAND_EXECUTION]: The skill executes
olclicommands to interact with Overleaf's API, allowing for project listing, pulling/pushing files, and triggering remote compilations. - [PROMPT_INJECTION]: Indirectly ingests external LaTeX source code and project data from the Overleaf platform, which serves as a potential surface for indirect prompt injection.
- Ingestion points: The
olcli pull,olcli download, andolcli synccommands retrieve external project files (SKILL.md). - Boundary markers: No explicit delimiters are used to wrap the untrusted content when it is processed by the agent.
- Capability inventory: The skill has the ability to execute CLI commands, modify the file system, and perform network requests via the
olclitool (SKILL.md). - Sanitization: No sanitization or filtering is performed on the content of the LaTeX files pulled from the remote repository.
Audit Metadata