research-manager

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a robust organizational workflow for research projects using standard task management principles.
  • State Persistence: The use of project.json and research-tasks.md within a designated workspace directory (research/{slug}/) facilitates session continuity for the research project. This is a legitimate functional requirement for managing long-running tasks and does not involve malicious persistence mechanisms.
  • Controlled Ingestion: The workflow explicitly scopes ingestion to user requests and output from another internal skill (research-methodology), maintaining a standard orchestration pattern.
  • Tool Usage: The requested tools (Read, Write, Task, WebSearch) are consistent with the skill's stated purpose of managing research and documenting findings.
  • No Obfuscation: The instructions and JSON schema are clear, human-readable, and free from hidden or encoded content.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 05:29 PM
Security Audit — agent-trust-hub — research-manager