Submit Feedback
Read shared execution guidance once per task before using tools. It covers existing authorization, model discovery, optional cost checks, imports, and recovery.
Turn the user's report into one concise, useful submit_feedback call. Feedback is a support channel; it is not a generation tool and does not guarantee a reply or roadmap commitment.
When to use
- A tool failed, returned the wrong state, or behaved inconsistently.
- An image, video, or voice result had a repeatable quality problem.
- A workflow or instruction was confusing.
- The user asks for a missing feature, integration, format, or model.
- The user explicitly asks to send praise or product feedback.
Do not submit ordinary dissatisfaction without enough context, silently report generation failures, or use this skill when the user only wants help revising media. If you observed the issue rather than receiving an explicit request, offer to report it and wait for approval.
Build the report
Include the user's intended outcome, what happened, expected behavior, useful reproduction details, model or tool involved, and practical impact. Exclude secrets, credentials, unnecessary personal information, and unsupported guesses.
Map the report to the exact schema:
category:bug,missing_feature,confusing,praise, orothersource:userwhen explicitly requested or relayed;agentfor an agent-observed product issuemessage: the concise reportattemptedTask: the task the user was trying to complete, when relevanttoolName: the exact MCP tool involved, when known
Submit and continue
After the user asks or approves, call submit_feedback once per distinct issue. Confirm what category was sent without promising a response. If the user still needs help, continue with a safe workaround or corrected workflow after submitting.