chapter-lead-writer

Warn

Audited by Gen Agent Trust Hub on Feb 17, 2026

Risk Level: MEDIUMPROMPT_INJECTIONNO_CODE
Full Analysis
  • Obfuscation (MEDIUM): Hex escapes (\x27) were detected in the description and workflow steps. While these decode to standard apostrophes ('), the use of hex encoding for punctuation matches Category 3 patterns and requires review for potential obfuscation of intent.
  • Indirect Prompt Injection (LOW): The skill is vulnerable to indirect prompt injection because it ingests external content to influence agent output.
  • Ingestion points: Processes data from outline/outline.yml, outline/chapter_briefs.jsonl, outline/writer_context_packs.jsonl, and citations/ref.bib.
  • Boundary markers: Absent. The instructions do not provide the agent with specific delimiters or warnings to treat external file content as non-executable data.
  • Capability inventory: The agent has the capability to write to the local file system at sections/S<sec_id>_lead.md.
  • Sanitization: The skill performs basic validation for citation keys against a reference bibliography, but does not sanitize the primary prose content ingested from JSONL/YAML files.
Audit Metadata
Risk Level
MEDIUM
Analyzed
Feb 17, 2026, 06:06 PM