research-proposal

Pass

Audited by Gen Agent Trust Hub on Oct 11, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill fetches scholarly metadata and research papers from well-known and trusted academic services, including the Crossref REST API, arXiv, and PubMed, to verify citations and extract research findings for proposals.
  • [COMMAND_EXECUTION]: The skill uses the Bash tool to execute grep for automated validation of citation placeholders and pandoc for converting markdown proposals into standard academic formats like Word or PDF.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data from research abstracts and documents retrieved from the web or the user's Zotero library. Ingestion points: External metadata and full-text documents are retrieved via WebFetch and Zotero MCP tools (e.g., mcp__zotero__zotero_get_item_fulltext). Boundary markers: The skill lacks explicit text delimiters for external data but implements a 'Citation Integrity Protocol' to ensure accuracy. Capability inventory: The skill possesses Write, Edit, and Bash capabilities for document processing and integrity checks. Sanitization: There is no explicit sanitization for paper content, but claims are manually and automatically cross-verified against source metadata.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 11, 2026, 03:00 AM