ideabrowser-openclaw-flow
Pass
Audited by Gen Agent Trust Hub on Sep 3, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes external command-line utilities
playwright-cliandopenclawto interact with browser sessions and retrieve web content. - Evidence: Found in
scripts/crawl_idea_subpages.sh,scripts/crawl_idea_subpages_playwright.sh, andscripts/extract_idea_fields.sh. - [INDIRECT_PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection as it ingests untrusted content from external web pages to generate implementation plans.
- Ingestion points: The
snapshotcommands inscripts/crawl_idea_subpages.shandscripts/crawl_idea_subpages_playwright.shread content fromideabrowser.com. - Boundary markers:
SKILL.mddefines a 'safe-by-default' boundary, instructing the agent not to write files or deploy unless explicitly asked by the user. - Capability inventory: The agent can execute browser commands, list tabs, and generate markdown documentation based on scraped data.
- Sanitization: The helper scripts perform regex-based field extraction but do not implement rigorous sanitization of the full page snapshots before they are processed by the agent.
Audit Metadata