replicate-media-provider
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes the
renkucommand-line interface to interact with the media generation provider. It executes commands to list available models, retrieve specific model guidance, and fetch input schemas. These operations are restricted to internal platform tools and local project references. - [INDIRECT_PROMPT_INJECTION]: The skill incorporates external data from user instructions and personal model guidance files to author requests.
- Ingestion points: External data enters the context through user directions and local files (
personalGuidePath) retrieved via the CLI. - Boundary markers: The skill does not define specific delimiters to isolate potentially untrusted input within the authored requests.
- Capability inventory: The skill has the capability to write files to the project's
tmp/directory and initiate media generation workflows through the authorizedrenkuengine. - Sanitization: The skill relies on native validation and schema checks provided by the execution engine; it does not specify additional sanitization for text ingested from guidance files.
Audit Metadata