prioritize-context-data

Pass

Audited by Gen Agent Trust Hub on Jul 8, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The instructions establish a workflow that relies on untrusted data from context attachments (e.g., .xlsx, .pdf, .csv). This creates a surface for indirect prompt injection, where an attacker could embed malicious instructions within a data file that the agent then processes as a primary source of truth.
  • Ingestion points: Reading external files provided in the task context or attachment panel (SKILL.md, Workflow Step 1).
  • Boundary markers: Absent; the skill does not instruct the agent to use delimiters or disregard instructions found within the data.
  • Capability inventory: Mention of read_file, pandas, pdf_reader, and web search capabilities (SKILL.md, Workflow Steps 3 & 5).
  • Sanitization: Absent; no instructions are provided to sanitize, escape, or validate the content of the ingested files.
Audit Metadata
Risk Level
SAFE
Analyzed
Jul 8, 2026, 09:00 PM
Security Audit — agent-trust-hub — prioritize-context-data