software-directory-lead-scraper
Warn
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: MEDIUMREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill contains instructions to automatically install the 'BrowserAct' CLI and associated skills if they are missing. It explicitly tells the agent to perform this installation "without asking the user to confirm" and to "not request separate approval for dependency installation." This encourages the agent to execute installation scripts autonomously, bypassing standard security oversight.
- [EXTERNAL_DOWNLOADS]: The skill relies on a third-party service located at browseract.com and its associated GitHub repository for its core functionality. It directs the agent to navigate to these external URLs to fetch software dependencies.
- [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests data from untrusted external sources (B2B directories and marketplaces).
- Ingestion points: Web content processed via the browser-act tool (SKILL.md).
- Boundary markers: No markers or delimiters are specified to separate agent instructions from the scraped data.
- Capability inventory: The agent has access to external browsing tools and local storage in the workspaces directory.
- Sanitization: The instructions do not specify any validation or filtering of the external content before it is processed.
Audit Metadata