analyze

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: No malicious code, obfuscation, or unauthorized access patterns were found. The skill operates as a high-level orchestrator for planning and documentation within a local workspace.
  • [COMMAND_EXECUTION]: The skill does not execute shell commands or spawn subprocesses. Its operations are limited to coordinating internal AI tasks and generating local markdown files.
  • [DATA_EXFILTRATION]: Analysis of the skill instructions confirms the absence of network-enabled tools or operations. Data is processed locally and saved to a user-specified artifact folder.
  • [PROMPT_INJECTION]: The skill ingests untrusted data from user requests and local codebases (Ingestion points: Step 1 input, Step 2 codebase scan in SKILL.md). While no explicit boundary markers or sanitization routines are defined, the skill's capabilities are restricted to writing markdown files (Capability inventory: artifact generation in references/formats.md), and it enforces mandatory human review gates after every step of the pipeline.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 10:09 PM