lokalbot-cli
Pass
Audited by Gen Agent Trust Hub on Aug 21, 2026
Risk Level: SAFECOMMAND_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill invokes a local binary located at
/Applications/LokalBot.app/Contents/Helpers/lokalbot-clito perform data retrieval, search, and library management. - [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted external data (meeting transcripts, summaries, and screen OCR) which can be manipulated by an attacker to include hidden instructions targeting the agent.
- Ingestion points: Meeting metadata, transcripts, and screen activity are ingested through
lokalbot-cli get,search, and various MCP tools likeget_timelineandsearch_screen(SKILL.md). - Boundary markers: The skill lacks explicit delimiters or specific instructions to ignore potentially malicious commands embedded within the meeting content or screen text.
- Capability inventory: The agent has the ability to execute shell commands using the
lokalbot-clibinary. - Sanitization: There is no evidence of content sanitization, escaping, or validation applied to the meeting data before it is presented to the agent.
Audit Metadata