alterlab-akademik-tesvik
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes user-provided activity lists in JSON format to calculate scores. This data ingestion path constitutes a surface for indirect prompt injection where crafted input could attempt to influence the agent.
- Ingestion points: The script
scripts/tesvik_score.pyreads user-supplied JSON data from a file path or standard input provided at runtime. - Boundary markers: The skill lacks explicit delimiters or instructions to ignore embedded commands within the activity list data.
- Capability inventory: The skill is authorized to use the
Bashtool to execute a Python script which processes the ingested data. - Sanitization: The Python script implements input validation by normalizing keys using ASCII-folding and verifying them against a fixed whitelist of allowed activity types. It also enforces strict type casting of values to floats for all calculations.
Audit Metadata