claude-code-system-prompts

Fail

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: HIGHEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONREMOTE_CODE_EXECUTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill directs users to clone a repository from https://github.com/Piebald-AI/claude-code-system-prompts.git and install the tweakcc package from NPM. Neither source is identified as a trusted or well-known service in this context.
  • [COMMAND_EXECUTION]: The skill provides instructions for running high-risk commands such as npm install -g tweakcc and tweakcc apply. The latter is specifically designed to patch the installed Claude Code binary and its system prompts, which modifies the core logic of the agent environment.
  • [REMOTE_CODE_EXECUTION]: By encouraging the installation and execution of the tweakcc utility, the skill facilitates the execution of code from an external third-party whose integrity and safety cannot be verified.
  • [DYNAMIC_EXECUTION]: The skill includes JavaScript examples that perform dynamic file access using fs.readFileSync with computed paths and describes methods for extracting prompts from compiled binaries at runtime.
  • [DATA_EXPOSURE]: Documentation within the skill highlights sensitive configuration paths like ~/.claude-code and critical environment variables such as ANTHROPIC_API_KEY.
Recommendations
  • AI detected serious security threats
Audit Metadata
Risk Level
HIGH
Analyzed
May 16, 2026, 11:20 PM
Security Audit — agent-trust-hub — claude-code-system-prompts