audioshake-stem-separation
Pass
Audited by Gen Agent Trust Hub on Sep 20, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the ingestion of untrusted data from external sources, which can be exploited for indirect prompt injection if processed outputs are used by an agent to drive subsequent logic.
- Ingestion points: The skill utilizes the
urlandtranscriptUrlfields (as seen inSKILL.md) to pull content from remote HTTPS endpoints, alongsideassetIdandtranscriptAssetIdfor previously uploaded content. - Boundary markers: There are no explicit instructions or delimiters defined to separate untrusted external content from the agent's core instructions.
- Capability inventory: The skill enables network operations to the AudioShake API (
api.audioshake.ai) and ingestion of text-based outputs (transcripts and JSON metadata) into the agent context. - Sanitization: The provided documentation does not specify any sanitization or validation protocols for the content of the audio or the associated transcript data processed by the models.
Audit Metadata