book-author-orchestrator

Pass

Audited by Gen Agent Trust Hub on Jul 31, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted academic sources (RPS, research papers, thesis notes), which creates an attack surface for indirect prompt injection.
  • Ingestion points: Processes user-provided academic documents such as RPS and research papers as defined in SKILL.md.
  • Boundary markers: Lacks explicit delimiters in the instruction logic, but utilizes a human-in-the-loop review process via hitl_book_checklist.md.
  • Capability inventory: Executes local Python scripts (validate_references.py, etc.) and performs network lookups.
  • Sanitization: Employs a book-security-privacy-checker to identify and redact sensitive data at project intake and finalization.
  • [COMMAND_EXECUTION]: Executes local Python helper scripts (e.g., parse_rps.py, validate_references.py) stored in the .codex/skills/scripts/ directory for document analysis and knowledge graph indexing.
  • [EXTERNAL_DOWNLOADS]: Includes functionality to monitor for and download updates from the official vendor repository at https://github.com/lensetek/Book-Author-Agent-Skills.
  • [DATA_EXFILTRATION]: Interacts with well-known academic services such as Crossref, arXiv, and PubMed/NCBI to search for scholarly references and metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 31, 2026, 02:04 PM
Security Audit — agent-trust-hub — book-author-orchestrator