competitive-intel
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the Bright Data CLI (bdata) from the vendor's official domain cli.brightdata.com.\n- [REMOTE_CODE_EXECUTION]: The installation process involves piping a remote shell script directly to bash:
curl -fsSL https://cli.brightdata.com/install.sh | bash. This is the vendor's official distribution method.\n- [COMMAND_EXECUTION]: The skill frequently invokes the bdata command-line tool and uses shell utilities like jq and xargs to process scraped data programmatically within SKILL.md and its references.\n- [INDIRECT_PROMPT_INJECTION]:\n - Ingestion points: The skill retrieves data from external websites using bdata scrape and bdata pipelines across all analysis modules in SKILL.md.\n
- Boundary markers: Instructions in SKILL.md advise the agent to separate facts from analysis and cite source URLs, but do not provide explicit delimiters to isolate untrusted web content from instructions.\n
- Capability inventory: The skill possesses command execution capabilities via the bdata CLI and shell pipes for data processing.\n
- Sanitization: No specific sanitization or filtering of external HTML or Markdown content is described before the data is processed by the agent.
Audit Metadata