deploy-openclaw-bot-for-me
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches deployment instructions from the author's official GitHub Pages repository at
https://vinta.github.io/dear-ai/deploy-openclaw-bot-for-me.md. - [REMOTE_CODE_EXECUTION]: The skill body instructs the agent to follow instructions fetched from a remote URL. As the URL corresponds to the author's own domain, this is documented as standard vendor functionality.
- [INDIRECT_PROMPT_INJECTION]: The skill executes instructions from a remote source, which introduces a potential attack surface for indirect prompt injection.
- Ingestion points: The agent is directed to fetch and follow content from a remote URL.
- Boundary markers: No explicit delimiters or warnings are present to isolate the remote content from the local context.
- Capability inventory: The skill is configured with access to file system tools (
Read,Write,Edit), network fetching (WebFetch), and terminal commands (Bash). - Sanitization: No integrity checks or sanitization steps are defined for the remote content.
- Mitigation: Wrap external content in delimiters with explicit instructions to ignore embedded commands.
- [COMMAND_EXECUTION]: The skill requires access to
ssh,ssh-keygen, andchmodcommands to manage authentication and deploy the bot to a remote server.
Audit Metadata