yandex-metrika
Pass
Audited by Gen Agent Trust Hub on Sep 16, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDATA_EXFILTRATIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes data from the Yandex Metrika API which may include untrusted input such as search phrases or UTM parameters. The skill mitigates this by limiting output to the first 30 lines, using structured data files (CSV/TSV) to avoid context pollution, and employing a json_escape function in scripts/common.sh to sanitize data.
- [COMMAND_EXECUTION]: The skill executes local shell scripts and Python helpers. It implements strong security controls, including verifying that the .env file has restricted permissions (chmod 600) and using private temporary files to store OAuth headers, ensuring tokens are not exposed in process lists.
- [DATA_EXFILTRATION]: Network activity is directed exclusively to the well-known Yandex Metrika API at api-metrika.yandex.net. The skill correctly handles sensitive account identifiers and tokens, ensuring they are only transmitted to official service endpoints.
- [DYNAMIC_EXECUTION]: The skill invokes a local Python helper script (scripts/management_json.py) to handle complex JSON parsing. This is a secure practice to avoid fragile shell-based parsing and is restricted to the skill's own distributed code.
Audit Metadata