clawhub-skill-mng
Warn
Audited by Gen Agent Trust Hub on Sep 14, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSPERSISTENCECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill requires the installation of the 'clawhub' package from npm if it is not already present on the system.
- Evidence: SKILL.md contains instructions for
npm install -g clawhubandnpm install --prefix ~/.copilot-shell/bin clawhub. - [EXTERNAL_DOWNLOADS]: All operations are configured to use a non-standard third-party registry mirror rather than the official registry.
- Evidence: Multiple commands in SKILL.md include the flag
--registry https://cn.clawhub-mirror.com. - [PERSISTENCE]: The skill instructs the agent to modify the user's shell configuration files to ensure the installed tool persists in the system path.
- Evidence: Instructions to append export commands to
~/.bashrcor~/.zshrcfollowed bysourcecommands in SKILL.md. - [INDIRECT_PROMPT_INJECTION]: The skill acts as an interface for searching, inspecting, and installing other skills from an external marketplace, creating a surface where malicious instructions in external skill files could influence the agent.
- Ingestion points:
clawhub search,clawhub explore, andclawhub inspect <slug> --file SKILL.md(reading external skill instructions). - Boundary markers: No delimiters or ignore instructions are specified for the external output.
- Capability inventory: The skill has shell command execution (npm, clawhub) and file writing capabilities (modifying ~/.bashrc).
- Sanitization: No sanitization or validation of the external marketplace content is described.
Audit Metadata