Douyin Content Assistant
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
No arguments | |||
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
This server publishes no instructions, or was last inspected before Glama recorded them.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": false
} |
| prompts | {
"listChanged": false
} |
| resources | {
"subscribe": false,
"listChanged": false
} |
| experimental | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| check_creatorA | Start a local creator-name/ID/profile check and return check_id immediately. Poll get_report. No scheduler is created; only a complete reviewed report advances delivery history. |
| get_reportB | Get live status, newest link, sampled like/comment/collect/share counts and raw/reviewed file paths. done+complete means raw readiness; delivery_complete additionally requires host-AI review. |
| get_transcriptB | Read ordered original ASR segments with indices and source_sha256. Follow next_offset until null before reviewing. Never summarize the title as speech. |
| submit_reviewA | Save a semantic review supplied by the current host AI, never call another LLM. Requires source_sha256, edits[{index,before,after,reason,evidence,evidence_level,confidence}], pending_review[{index,issue}], overview{summary,sections[{start,end,title,summary}],limitations}. Leaves originals immutable; uncertain text remains pending. |
| start_loginA | Prepare a dedicated local Douyin login QR. User scans normally. Poll login_status for image path; no other browser profiles or cookie contents are read. |
| login_statusA | Return QR path and dedicated-session status only. Saved state can expire; never return browser credentials. |
| resolve_creator_from_videosA | Fallback: host web search supplies 1-5 observed official Douyin video URLs; verify live author matches exact requested identity. Search index is discovery only, never proof of newest work or counts. Stops on captcha/ambiguity. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 7 tools
Most tools have clearly distinct roles in a staged workflow: login, creator check, report polling, transcript reading, review submission, and fallback creator resolution. The main possible confusion is between check_creator and resolve_creator_from_videos, since both concern creator identity, but the descriptions frame the latter as a fallback and the former as the primary path.
The set mostly follows a predictable snake_case verb_noun pattern (check_creator, get_report, get_transcript, submit_review, start_login, resolve_creator_from_videos). The exception is login_status, which is a noun/status phrase rather than a clear verb_noun action, but overall the naming remains readable and consistent.
Seven tools is well-scoped for a focused Douyin content assistant pipeline. Each tool covers a distinct step in the workflow without obvious bloat or missing obvious counterparts.
The tool surface covers the core lifecycle: authentication, creator checking, report retrieval, transcript access, review submission, and a fallback creator resolution path. Minor gaps exist, such as no explicit tool for accessing delivery history or past reports despite those states being referenced, but the main workflows are represented.