channel-quality-analysis
Pass
Audited by Gen Agent Trust Hub on Sep 8, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [NO_CODE]: The skill consists entirely of natural language instructions and YAML configuration. There are no scripts (Python, JavaScript), executables, or binary files included in the skill package.
- [COMMAND_EXECUTION]: The skill provides instructions for the agent to use an external tool (
ae-cli) to perform data queries. These commands are parameter-controlled and follow a strict delegation protocol where the skill specifies the query intent and time range, and the tool handles the execution. - [INDIRECT_PROMPT_INJECTION]: The skill includes a 'Relay Summary Recognition' feature designed to process context from previous sessions. While this is an ingestion point for external data, the skill uses a structured template and multi-stage verification to ensure the data is relevant to channel analysis, minimizing the risk of prompt injection from data sources.
- [DATA_EXFILTRATION]: All data operations involve querying internal project metrics using
ae-cli. The skill explicitly forbids proactively mentioning specific client names or citing internal data in archived documents, and does not perform network operations to non-whitelisted or suspicious external domains.
Audit Metadata