flow-paper-lifecycle
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill functions as a high-level workflow manager and router for academic research processes. It does not implement sensitive functionalities directly but coordinates transitions between specialized research-focused skills.
- [PROMPT_INJECTION]: Analysis of the orchestrator instructions reveals no attempts to override agent behavior, bypass safety guardrails, or perform system prompt extraction. The language is purely procedural and focused on the research lifecycle.
- [DATA_EXFILTRATION]: No network operations, credential harvesting, or unauthorized file access patterns were detected. The skill's data interaction is limited to reading its own included reference documents and processing user-provided research context.
- [REMOTE_CODE_EXECUTION]: The skill does not perform any remote code downloads or dynamic execution of untrusted scripts. References to research codebases and training recipes are managed through delegation to other specialized skills, and this orchestrator contains no code execution primitives.
- [COMMAND_EXECUTION]: No shell commands, system calls, or use of the dynamic context injection syntax (
!command) were found in any of the skill files. - [SAFE]: The skill implements a robust internal logic system using evidence flags (EXPERT, SUPPORTED, CONTESTED, REFUTED) to maintain epistemic integrity in its advice, representing a best-practice for research-oriented agents.
Audit Metadata