conductor-methodology

Pass

Audited by Gen Agent Trust Hub on Feb 21, 2026

Risk Level: SAFENO_CODE
Full Analysis
  • Indirect Prompt Injection (SAFE): The skill utilizes a methodology where the agent reads and updates local repository files within the conductor/ directory to manage project context. \n
  • Ingestion points: Reads from conductor/product.md, tech-stack.md, and track files. \n
  • Boundary markers: The framework mandates explicit user confirmation before starting implementation and at phase boundaries, which serves as an effective control. \n
  • Capability inventory: Standard agent tools for file system access (Read, Write, Edit) and repository management (Git). \n
  • Sanitization: Structured markdown plans and human verification of specs mitigate the risk of accidental obedience to instructions embedded in data. \n- Remote Code Execution (SAFE): There are no patterns involving the download or execution of remote scripts or packages from external URLs. \n- Data Exposure & Exfiltration (SAFE): No sensitive file paths, hardcoded credentials, or non-whitelisted network operations were identified. The skill only interacts with internal project-specific documentation. \n- Privilege Escalation (SAFE): The instructions do not involve any operations requiring elevated permissions, such as sudo, or modifications to system-level configurations. \n- Command Execution (SAFE): Tools like bash and git are described only in the context of standard development tasks (e.g., commits, tests) following the documented workflow.
Audit Metadata
Risk Level
SAFE
Analyzed
Feb 21, 2026, 02:23 PM