azure-ai-transcription-py
Warn
Audited by Gen Agent Trust Hub on Aug 9, 2026
Risk Level: MEDIUMEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill instructs the installation of
azure-ai-transcription, which is not an official Microsoft package (official versions includeazure-ai-speechorazure-cognitiveservices-speech). Using non-standard package names from public registries carries a risk of typosquatting or supply chain compromise. - [COMMAND_EXECUTION]: The provided instructions require the execution of shell commands (
pip install) to add external dependencies to the environment. - [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection attacks because it processes untrusted data from external audio sources. 1. Ingestion points: include the
content_urlsparameter in thebegin_transcriptionmethod and local audio files insend_audio_file. 2. Boundary markers: are absent; there are no instructions defined to prevent the agent from following commands embedded in the transcribed text. 3. Capability inventory: includes the use of Python and network-enabled SDKs. 4. Sanitization: is absent; no validation or filtering is applied to the transcribed output before it is returned to the agent context.
Audit Metadata