get_debug_feedbacks
Pull user debug feedback records for the selected app, download screenshots and logs, and generate AI-ready debug context files.
Instructions
[H5 Debug Workflow]
Pull user debug feedback records for the selected app, download artifacts (screenshots/logs),
and generate AI-ready debug context files.
**PREREQUISITE: An app MUST be selected first.**
Before calling this tool, ALWAYS call get_current_app_info to verify an app is selected.
If not selected, guide user through:
1) Call list_developers_and_apps to show available apps
2) Show list to user and ASK them to choose
3) Call select_app with user's choice
4) Then call this tool
**DEFAULT BEHAVIOR:**
- fetch_and_mark_processed defaults to true
- download_assets defaults to true
- downloaded files are saved under logs/feed_back/feedback_{id}/
**CALLING POLICY FOR AGENTS:**
- If user says "拉取/查看反馈" without explicit read-only intent, DO NOT pass fetch_and_mark_processed.
Let default behavior (true) apply.
- ONLY pass fetch_and_mark_processed=false when user explicitly requests read-only
behavior (e.g. "只查看，不标记处理").
- When user provides a moment_id (or a TapTap moment URL), pass it via moment_id param to fetch a single feedback.
In this case, fetch_and_mark_processed is automatically forced to false (read-only single lookup).
Input Schema
TableJSON Schema
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | How many feedback records to pull (1-10, default 3). | |
| status | No | Feedback status filter: - 0: all - 1: unprocessed - 2: processed When fetch_and_mark_processed=true, status filter will be ignored by server. | |
| moment_id | No | Optional. Filter to a single feedback by moment_id. Accepts either a pure numeric ID (e.g. "795659996946762795") or a full TapTap moment URL (e.g. "https://www.taptap.cn/moment/795659996946762795") — the handler will extract the numeric ID. When provided, server returns only the matching record, and fetch_and_mark_processed is forced to false (read-only single lookup). MUST be passed as a string to preserve precision for large IDs. | |
| download_assets | No | Whether to download feedback JSON/screenshot/log files to local workspace. Default true. | |
| fetch_and_mark_processed | No | Pull unprocessed records and mark them processed on server. Default true. Set false only for explicit read-only requests. When true, status is ignored. |