mmk-plaud
Pass
Audited by Gen Agent Trust Hub on Jun 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill uses the
mmkCLI tool to perform various operations including listing recordings, downloading audio, renaming files, and executing administrative API passthroughs. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it ingests and processes untrusted external data.
- Ingestion points: Transcription text and recording summaries are retrieved from the Plaud API via the
mmk plaud file listandmmk plaud file transsummcommands (SKILL.md). - Boundary markers: The instructions do not provide delimiters or specific warnings to the agent to treat the fetched transcript content as data rather than instructions.
- Capability inventory: The agent has the ability to execute shell commands via
Bash(mmk *), which includes file manipulation and administrative API access (SKILL.md). - Sanitization: There is no evidence of sanitization or validation performed on the transcription data before it is presented to the agent context.
Audit Metadata