sdr-call-coaching
Pass
Audited by Gen Agent Trust Hub on Oct 8, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external data that could potentially contain malicious instructions if a prospect or recording were manipulated to influence the agent's behavior.
- Ingestion points: Call transcripts and summaries are retrieved via the Clari
clari_get_calland NookslistCallstools. - Boundary markers: The skill uses structured scoring rubrics which helps constrain the AI's output, though it does not explicitly define delimiters for the raw transcript text.
- Capability inventory: The agent has the ability to send Slack messages, create Gmail drafts, and write local HTML/JSON files.
- Sanitization: While the skill mentions summarizing large transcripts, it does not detail specific methods for escaping or filtering untrusted text within those transcripts.
- [COMMAND_EXECUTION]: The skill uses the bash
datecommand to dynamically calculate the current week's reporting window (Monday to Friday). This is a safe and standard use of shell commands for date arithmetic.
Audit Metadata