wiki-ingest

Pass

Audited by Gen Agent Trust Hub on Jun 23, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: Indirect Prompt Injection Surface detected in the document ingestion workflow.
  • Ingestion points: The skill reads external files (PDFs, markdown, transcripts) and vault notes using the Read and obsidian_read_note tools.
  • Boundary markers: Absent. The workflow does not utilize delimiters or specific instructions to help the agent distinguish between source content and its own processing logic.
  • Capability inventory: The skill possesses significant capabilities including vault searching (obsidian_rag_query), note creation (obsidian_create_note), and section modification/frontmatter updates (obsidian_insert_at_heading, obsidian_replace_section, obsidian_update_frontmatter).
  • Sanitization: Absent. There is no evidence of filtering or escaping content extracted from external sources before it is interpolated into the extraction logic or written to the vault.
  • Mitigation: The risk is significantly reduced by Step 5, which mandates presenting an 'Ingest Plan' and waiting for explicit user confirmation before any writes or merges occur.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 23, 2026, 10:34 AM
Security Audit — agent-trust-hub — wiki-ingest