paper-mcp

Pass

Audited by Gen Agent Trust Hub on Mar 28, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it retrieves and processes untrusted content from external academic papers to generate structured outputs and implementation plans.
  • Ingestion points: Untrusted data enters the context via tool calls that fetch full text or section-level content as described in SKILL.md (Required flow step 2) and the tool mapping checklist in references/paper-tools-and-prompts.md.
  • Boundary markers: The prompt templates and instructions lack explicit delimiters or mandatory "ignore embedded instructions" warnings for the fetched paper content.
  • Capability inventory: The skill allows the agent to fetch metadata, sections, figures, and citations, and then translate that content into implementation-ready notes and code-facing plans.
  • Sanitization: There is no mention of sanitizing or validating the fetched text for malicious instructions before it is processed by the LLM.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 28, 2026, 02:15 AM