write-protocol

Pass

Audited by Gen Agent Trust Hub on Sep 17, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies a potential surface for indirect prompt injection as it is designed to ingest and verbatim embed content from external project artifacts generated by other skills in the research pipeline.
  • Ingestion points: The skill reads content from protocol/sample_size_justification.md and outputs from design-study, search-lit, and define-variables.
  • Boundary markers: Absent; the skill is instructed to incorporate inputs directly without delimiters or explicit instructions to ignore embedded commands within those inputs.
  • Capability inventory: The skill has access to tools including Bash, Write, Edit, Read, Grep, and Glob.
  • Sanitization: Absent; the workflow relies entirely on the integrity of the data from preceding stages of the research pipeline.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 17, 2026, 02:00 AM
Security Audit — agent-trust-hub — write-protocol