paper-spine-rewrite

Pass

Audited by Gen Agent Trust Hub on Jun 13, 2026

Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill executes local Python scripts (scripts/integrity_audit.py, scripts/structured_review.py, and scripts/revision_audit.py) to automate manuscript integrity and logic auditing. These commands are restricted to local file operations within the project directory and are used for workflow validation.
  • [PROMPT_INJECTION]: The skill processes user-provided manuscript drafts which creates an indirect prompt injection surface when this data is used to generate rationale matrices or passed to sub-agents for review. 1. Ingestion points: User drafts extracted from the draft_path or conversation history. 2. Boundary markers: The skill employs structured tables (Original Logic Map, Writing Rationale Matrix) as logical boundaries to separate argument structure from underlying prose, providing structural constraints. 3. Capability inventory: Execution of local Python scripts and file-system writing for generated audit reports. 4. Sanitization: No specific input sanitization or validation of the manuscript text was observed in the provided instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 13, 2026, 03:36 AM
Security Audit — agent-trust-hub — paper-spine-rewrite