3scribe

Pass

Audited by Gen Agent Trust Hub on Mar 4, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill instructions in SKILL.md utilize the membrane CLI to perform administrative tasks and data operations, such as running actions and API requests.
  • [EXTERNAL_DOWNLOADS]: The documentation in SKILL.md recommends installing the @membranehq/cli global package, which is an official tool provided by the vendor to facilitate the integration.
  • [PROMPT_INJECTION]: The skill processes transcription and job data from 3Scribe, creating an indirect prompt injection surface. (1) Ingestion points: Output from membrane action run and membrane request commands as described in SKILL.md. (2) Boundary markers: Absent; the instructions do not define delimiters or warnings for the agent to ignore instructions embedded in transcription data. (3) Capability inventory: The agent can execute membrane CLI commands to interact with the network and execute actions. (4) Sanitization: Absent; the skill does not specify any validation or filtering of the content retrieved from the 3Scribe service.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 4, 2026, 08:56 AM