salvage

Pass

Audited by Gen Agent Trust Hub on Aug 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill manages session persistence by reading and writing to files within a localized hidden directory (.oh/). These operations are transparent, restricted to the project scope, and do not involve sensitive system configuration files or credentials.
  • [SAFE]: The skill integrates with Model Context Protocol (MCP) servers (Open Horizons, RNA) to query the local code corpus and log extracted insights. These tool invocations are standard for repository-native alignment and do not involve unauthorized network access.
  • [SAFE]: The skill instructs the agent to facilitate updates to project-wide documentation (e.g., AGENTS.md) with human-approved learnings. This collaborative workflow ensures user oversight and aligns with intended development practices.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 6, 2026, 04:54 PM
Security Audit — agent-trust-hub — salvage