show-me-your-work

Pass

Audited by Gen Agent Trust Hub on Nov 8, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from command outputs and external citations to record evidence in a decision trail.
  • Ingestion points: External command outputs and citations are logged into the evidence column of the TSV file (SKILL.md).
  • Boundary markers: The output is structured using TSV (Tab-Separated Values) columns.
  • Capability inventory: The skill involves writing and appending to local files within the project's state directory (.omc/ or OMC_STATE_DIR).
  • Sanitization: The skill explicitly implements a 'cell guard' that prefixes any cell starting with =, +, -, or @ with a single quote ('). This is a security measure designed to prevent formula injection attacks when the log file is opened in spreadsheet software like Excel or Google Sheets.
  • [EXTERNAL_DOWNLOADS]: The skill's metadata contains references to external GitHub repositories for attribution and credit (e.g., links to the original pstack plugin and its port). These are static documentation links and do not involve the automated downloading or execution of remote code during the skill's operation.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 8, 2026, 07:21 PM