tashkeel
Pass
Audited by Gen Agent Trust Hub on Nov 5, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: The skill utilizes the Farasa API (farasa-api.qcri.org) for Arabic linguistic processing. This is a well-known service developed by the Qatar Computing Research Institute (QCRI) and is considered a reputable source in the NLP community. It also installs the standard pyarabic library from PyPI.
- [INDIRECT_PROMPT_INJECTION]: The skill defines a workflow for processing user-supplied Arabic text, which represents a potential surface for indirect injection if the processed data is used to influence further agent actions.
- Ingestion points: User-provided Arabic text strings passed to curl or Python scripts.
- Boundary markers: None identified in the provided templates.
- Capability inventory: Network access via curl to a specific research API and local Python library execution.
- Sanitization: No explicit input sanitization is performed, though the nature of the task (diacritization) typically treats input as literal text.
Audit Metadata