awesome-claude-code-subagents
Fail
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: HIGHREMOTE_CODE_EXECUTIONEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [REMOTE_CODE_EXECUTION]: The skill documentation encourages users to download a shell script (
install-agents.sh) from a third-party GitHub repository and execute it locally usingchmod +xand direct execution. This pattern facilitates the execution of unverified remote code. - [EXTERNAL_DOWNLOADS]: Fetches installer scripts and agent configuration files (markdown prompts) from a non-standard external source (
https://raw.githubusercontent.com/VoltAgent/...). - [COMMAND_EXECUTION]: Instructs users to perform file system operations and execute shell scripts retrieved from the internet, which can be used as a vector for malicious system modifications.
- [INDIRECT_PROMPT_INJECTION]: The skill creates a vulnerability surface by importing 'subagent' prompts from an external repository.
- Ingestion points: Remote markdown files in the
VoltAgent/awesome-claude-code-subagentsrepository. - Boundary markers: None identified; the skill assumes all downloaded subagent content is safe to include in the agent context.
- Capability inventory: The instructions involve network access (
curl,git), file writes (cp,-o), and shell execution (bash). - Sanitization: No validation or sanitization of the remote agent instructions is performed before installation into the
~/.claude/agents/directory.
Recommendations
- AI detected serious security threats
Audit Metadata