kb-ingest
Pass
Audited by Gen Agent Trust Hub on Oct 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted document content from the Inbox or user-provided paths to perform classification, summarization, and linking. This creates a vulnerability where malicious instructions embedded in documents could influence the agent's actions. * Ingestion points: Full document content read from Inbox or specified paths (Step 1 in SKILL.md). * Boundary markers: Absent; there are no instructions to use delimiters or ignore instructions within the ingested data. * Capability inventory: The skill has permissions to read, write, move, and rename files, as well as execute Obsidian CLI commands. * Sanitization: Absent; document content is directly interpolated for summarization and metadata generation.
- [COMMAND_EXECUTION]: The skill utilizes the obsidian CLI tool (search, backlinks, tags) for vault management. These commands are executed based on data extracted from documents being processed, representing a command execution surface driven by untrusted input.
Audit Metadata