everything-claude-code-agent-harness
Warn
Audited by Gen Agent Trust Hub on May 16, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSREMOTE_CODE_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs users to clone a repository from 'github.com/affaan-m/everything-claude-code.git'.
- [REMOTE_CODE_EXECUTION]: The installation process involves running 'npm run install' and building a Rust control plane with 'cargo build' from the downloaded source code.
- [COMMAND_EXECUTION]: The skill utilizes lifecycle hooks (e.g., 'load-context.sh', 'save-summary.sh') that execute bash scripts to manage session state and context.
- [COMMAND_EXECUTION]: Instructions include modifying agent behavior by concatenating files into core configuration paths such as '
/.clinerules' and '/.cursorrules'. - [DATA_EXFILTRATION]: The system maintains detailed session history, token usage metrics, and embeddings within a local '.ecc/' directory. While no external exfiltration was observed, it processes and stores sensitive conversation data locally for its 'continuous learning' features.
- [PROMPT_INJECTION]: The 'Continuous Learning' feature (Category 8) extracts patterns from session logs into new skills, creating an ingestion point for potentially malicious instructions from untrusted data (Indirect Prompt Injection).
Audit Metadata