lov-human-writing

Pass

Audited by Gen Agent Trust Hub on Sep 8, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts, measure.py and profile_store.py, to analyze text and persist user preferences. These scripts utilize the Python standard library and do not perform network operations.
  • [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted manuscripts and drafts to identify stylistic patterns and suggest rewrites. This ingestion creates a potential surface for indirect prompt injection if the source text contains adversarial instructions meant to influence the agent's output.
  • Ingestion points: Drafts provided as file inputs or text strings in Step 1 of the SKILL.md workflow.
  • Boundary markers: The process follows a structured workflow involving an authorship ledger and a multi-dimensional discourse audit.
  • Capability inventory: The skill uses file reading, writing, and shell execution (via the Bash tool) to process local assets.
  • Sanitization: The primary stylistic audit is deterministic and regex-based, providing an objective layer before the agent performs rewriting tasks.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 8, 2026, 03:05 AM
Security Audit — agent-trust-hub — lov-human-writing