research-data-management

Pass

Audited by Gen Agent Trust Hub on Mar 23, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill functions as an indirect prompt injection surface by ingesting various project files (e.g., notebooks, configuration files, and metadata records) and using that data to generate reports.
  • Ingestion points: Processes contents of requirements.txt, package.json, pom.xml, ontology files (.owl, .ttl, .rdf), and lab notebooks (.ipynb, .Rmd) as specified in Phase 1.
  • Boundary markers: Lacks explicit delimiters or instructions to ignore embedded commands within the ingested research data.
  • Capability inventory: Includes file-write access to the project's docs/ directory for report generation and the ~/.claude/projects/ telemetry path for execution logging.
  • Sanitization: Does not implement sanitization or validation for content extracted from research assets before output generation.
  • [COMMAND_EXECUTION]: The skill executes filesystem write operations to create analysis reports and appends execution metadata to ~/.claude/projects/skill-telemetry.md for session tracking and self-evolution.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 23, 2026, 10:58 AM
Security Audit — agent-trust-hub — research-data-management