threads-mcp-cli MCP server
Commit 6. Nov. 2026Ermöglicht es AI Agents und CLI-Nutzern, Threads-Accounts zu verwalten, indem sie Beiträge posten, antworten, Insights analysieren und nach Inhalten suchen.
Auch bekannt als thenavidm/threads-mcp-cli
- Stars
- 4
- Tools
- 30
- Downloads, 30 Tage
- —
- Neueste Version
- —
- Letzter Commit
- 6. Nov. 2026
- Lizenz
- MIT
Was ist threads-mcp-cli MCP?
Ermöglicht es AI Agents und CLI-Nutzern, Threads-Accounts zu verwalten, indem sie Beiträge posten, antworten, Insights analysieren und nach Inhalten suchen. Er stellt 30 Tools bereit, etwa list_accounts, whoami, get_publishing_limit.
Zusammenfassung generiert aus dem README des Projekts
Über threads-mcp-cli MCP server
Dieses Tool ermöglicht es, Threads-Accounts über einen AI Agent oder das Terminal zu verwalten. Es bietet eine einheitliche Schnittstelle sowohl für CLI-Nutzer als auch für AI-Clients wie Claude Code, Cursor und Claude Desktop. Mit einer einzigen Installation erhältst du Zugriff auf 30 verschiedene Tools, die alles von einfachen Posts bis hin zu fortgeschrittenem Account-Management abdecken.
Nutzer interagieren mit dem Service entweder über ein CLI für Scripting und Cron-Jobs oder als MCP server für konversationelle KI. Das Setup erfordert eine Meta App ID und ein Secret, um ein 60-Tage-Token zu generieren, das sich automatisch erneuert. Da Threads eine eigene API nutzt, die getrennt von Instagram ist, ist ein dediziertes Token erforderlich.
Die Funktionen umfassen das Erstellen von Posts, Threads und Carousels sowie die Verwaltung von Antworten und Approval-Queues. Du kannst zudem tiefgehende Analysen durchführen, indem du Tools wie get_top_posts nutzt, um Inhalte basierend auf dem Engagement im Verhältnis zu den Views zu ranken, oder search-keyword verwendest, um nach spezifischen Themen zu suchen. Um die Sicherheit zu gewährleisten, erfordert jede Aktion, die Inhalte veröffentlicht, darauf antwortet oder löscht, eine explizite Bestätigung.
Was du damit machen kannst
- 1Content-Performance analysierenLass deinen AI Agent herausfinden, welche Posts am besten performt haben, indem er Engagement-Metriken mit den Gesamtaufrufe vergleicht.
- 2Social Media Posting automatisierenNutze die Kommandozeile, um Notizen in einen strukturierten Thread zu verwandeln oder ein Carousel mit Bildern zu veröffentlichen.
- 3Community-Interaktionen moderierenIdentifiziere, welche Antworten in deiner Konversation eine Rückmeldung verdienen, oder verstecke problematische Kommentare.
- 4Brand Mentions überwachenSuche nach spezifischen Keywords oder Topic-Tags, um zu sehen, was die Leute über einen Launch sagen.
Geschrieben aus dem README des Projekts. Jeder Use Case wurde gegen ein Zitat daraus geprüft.
threads-mcp-cli MCP tools 30
Erkannt aus dem READMElist_accountswhoamiget_publishing_limitrefresh_tokencreate_postcreate_threadcreate_carouselstage_postpublish_stagedget_container_statusquote_postrepost
| list_accounts | — | |
| whoami | — | |
| get_publishing_limit | — | |
| refresh_token | — | |
| create_post | — | |
| create_thread | — | |
| create_carousel | — | |
| stage_post | — | |
| publish_staged | — | |
| get_container_status | — | |
| quote_post | — | |
| repost | — |
Alle 30 Tools anzeigenWeniger Tools anzeigen
delete_postreply_toget_repliesget_conversationget_all_replieshide_replyget_pending_repliesmanage_pending_replyget_postsget_postget_post_insightsget_account_insightsget_follower_demographicsget_top_postssearch_keywordsearch_topic_taglookup_profilelist_allowlisted_countries
| delete_post | — | |
| reply_to | — | |
| get_replies | — | |
| get_conversation | — | |
| get_all_replies | — | |
| hide_reply | — | |
| get_pending_replies | — | |
| manage_pending_reply | — | |
| get_posts | — | |
| get_post | — | |
| get_post_insights | — | |
| get_account_insights | — | |
| get_follower_demographics | — | |
| get_top_posts | — | |
| search_keyword | — | |
| search_topic_tag | — | |
| lookup_profile | — | |
| list_allowlisted_countries | — |
Konfiguration
THREADS_ACCOUNTSTHREADS_ALLOW_DESTRUCTIVETHREADS_APP_IDTHREADS_APP_SECRETTHREADS_AUDIT_LOGTHREADS_DEFAULT_ACCOUNTTHREADS_HTTP_PORTTHREADS_HTTP_TOKENTHREADS_READ_ONLY
| Name | Beschreibung | Typ |
|---|---|---|
| THREADS_ACCOUNTS | — | |
| THREADS_ALLOW_DESTRUCTIVE | — | |
| THREADS_APP_ID | — | |
| THREADS_APP_SECRET | — | |
| THREADS_AUDIT_LOG | — | |
| THREADS_DEFAULT_ACCOUNT | — | |
| THREADS_HTTP_PORT | — | |
| THREADS_HTTP_TOKEN | — | |
| THREADS_READ_ONLY | — |
Automatisierte Checks
Geluften am 14. Nov. 2026Vollständiger Bericht →Noch nicht verifiziert. Ein Server benötigt einen Score von 80, keine kritischen Befunde und genug Daten für unsere Prüfung. Wie die Checks funktionieren · Was du für mehr Punkte verbessern kannst
- Source Repository verknüpft und erreichbarthenavidm/threads-mcp-cli is public and reachablegithub.com/thenavidm/threads-mcp-cli ↗Bestanden
- Paket nennt dieses Repository als QuelleNicht geprüft
- Build-Provenienz (Sigstore-Attestierung)Nicht geprüft
- Open Source Lizenz (OSI)MITspdx.org/licenses/MIT.html ↗Bestanden
- Commit oder Release im letzten JahrLast commit or release 2026-11-06github.com/thenavidm/threads-mcp-cli ↗Bestanden
- OpenSSF ScorecardNicht geprüft
- Package-Name ähnelt keinem populären PackageNicht geprüft
- Keine kritischen Advisories im Paket oder seinen AbhängigkeitenNicht geprüft
- Keine versteckten Anweisungen an das Modell30 tool descriptions and the README read; no override phrases, credential paths or hidden charactersBestanden
- Destruktive Tools sind markiertNicht geprüft
- Remote-Endpoint beantwortet den MCP-HandshakeNicht geprüft
README
Auf GitHub ansehenThreads MCP Server & CLI
Threads MCP server and CLI for Claude Code, Codex and AI agents. 30 tools for posting, chained threads, carousels, replies and reply approvals, insights, keyword search and profile discovery.
One install gives you both surfaces, the same tools under the same names, covering everything the app does and several things it cannot.
Threads has its own API, separate from Instagram's, so it needs its own token. One Meta app can carry both, with one app id and one testers list.
Publishing and deleting ask for confirmation. Everything else is a read.
One command to authorise, and the 60-day token refreshes itself from then on.
Built and maintained by Navid Moazzez. Built on Slipway, which turns one definition of each tool into the MCP server and the CLI.
Two ways to use it
Command line
threads-cli in your terminal, for scripting, cron, pipes, or a quick question
without opening anything:
threads-cli # every command, one line each
threads-cli whoami # which profile the token belongs to
threads-cli get-publishing-limit # how much quota is left today
threads-cli get-posts --limit 10 # your recent posts
threads-cli get-top-posts --limit 25 # ranked by engagement against views
threads-cli search-keyword "model context protocol"
threads-cli create-post --text "Shipped." --confirm
threads-cli list-accounts --json | jq -r '.accounts[].username'
threads-cli <command> --help # what any command takes
--confirm is the shell spelling of the confirmation that posting, replying and
deleting require. --json gives JSON, --compact puts it on one line, --agent
turns on all of the machine-readable defaults at once, and errors are JSON on
stderr whichever you pick.
threads-cli schema <command> prints the exact JSON Schema an MCP client
receives for that tool, which is how you can check the two surfaces really are
one thing.
Output and exit codes
Every command exits with a number a script can branch on, so nothing has to parse the message:
| Code | Means |
|---|---|
| 0 | It worked |
| 1 | An unexpected error, worth an issue |
| 2 | You typed it wrong, or the guard refused a write: a missing or bad flag, an unknown command, a command hidden by THREADS_READ_ONLY=1, a post without --confirm, Meta rejecting the arguments |
| 3 | Not found |
| 4 | The token was rejected, or lacks the permission |
| 5 | The Threads API failed or could not be reached |
| 7 | Rate limited, back off and retry |
| 10 | Nothing is configured yet, run threads-cli login |
threads-cli create-post --text "$BODY" --confirm --agent > /tmp/out.json
case $? in
0) ;;
2|4) echo "fix the command or the token, not retrying" >&2; exit 1 ;;
7) echo "rate limited, backing off" >&2 ;;
10) echo "no profile connected, run threads-cli login" >&2; exit 1 ;;
5) echo "Threads failed, will retry" >&2 ;;
*) echo "unexpected, see the error" >&2; exit 1 ;;
esac
MCP server, for your AI app
threads-mcp is what Claude Code, Claude Desktop, Cursor and the rest launch.
You never run it by hand:
claude mcp add threads -- npx -y @thenavidm/threads-mcp-cli
No credentials go in that line, because threads-cli login already stored the
token. Then just ask: "which of my posts this month actually worked, ranked by
engagement against views?"
Every other client is in section 2. Each post, reply, repost and delete waits for your approval in the client, as section 6 explains.
Which one
| Where you are | What you can reach |
|---|---|
| An agent that can run shell commands, like Claude Code or Cursor | Both. The CLI is the cheaper one: it costs nothing until you type it |
| claude.ai, the Claude Desktop chat tab, or a phone | The server only. There is no shell to run a command in |
| A terminal, a script, cron or CI | The CLI only. There is no MCP client in a shell |
They are the same program reading the same tool definitions, so anything one can do, the other can.
Features
Every tool is both a command and an MCP tool, with the same name. The command is the tool name with dashes.
| Capability | CLI command | MCP tool |
|---|---|---|
| Which profile, and the daily quota | threads-cli whoami / get-publishing-limit |
whoami / get_publishing_limit |
| List connected profiles | threads-cli list-accounts |
list_accounts |
| Post, thread or carousel | threads-cli create-post / create-thread / create-carousel |
create_post / create_thread / create_carousel |
| Stage a post, publish it later | threads-cli stage-post / publish-staged / get-container-status |
stage_post / publish_staged / get_container_status |
| Quote and repost | threads-cli quote-post / repost |
quote_post / repost |
| Delete one of your posts | threads-cli delete-post |
delete_post |
| Reply, and read replies | threads-cli reply-to / get-replies / get-conversation / get-all-replies |
reply_to / get_replies / get_conversation / get_all_replies |
| Hide a reply, or work the approval queue | threads-cli hide-reply / get-pending-replies / manage-pending-reply |
hide_reply / get_pending_replies / manage_pending_reply |
| Read your posts | threads-cli get-posts / get-post |
get_posts / get_post |
| Insights and what worked | threads-cli get-post-insights / get-account-insights / get-top-posts |
get_post_insights / get_account_insights / get_top_posts |
| Who follows you | threads-cli get-follower-demographics |
get_follower_demographics |
| Search posts and topic tags | threads-cli search-keyword / search-topic-tag |
search_keyword / search_topic_tag |
| Look up a public profile | threads-cli lookup-profile |
lookup_profile |
| Geo-gating countries | threads-cli list-allowlisted-countries |
list_allowlisted_countries |
| Keep the token alive | threads-cli refresh-token, or threads-cli refresh for every stored one |
refresh_token |
| Check your setup | threads-cli doctor |
not a tool |
All 30 with their arguments are in section 5.
Contents
| # | Section | What is in it |
|---|---|---|
| 1 | What you can ask it | Real prompts, not features |
| 2 | Install | Every client, copy and paste |
| 3 | Connect your account | The Meta app, in about ten minutes |
| 4 | What it costs to have connected | Measured in Claude Code, and how to spend less |
| 5 | Tools | All 30, with arguments |
| 6 | Writing safely | Why posting asks twice |
| 7 | Writing posts | Limits, media, threads, carousels |
| 8 | Reading posts | The output format, and why |
| 9 | Several profiles | Personal and brand, one server |
| 10 | Tokens | The 60-day clock, and how it is kept alive |
| 11 | How it works | Architecture |
| 12 | Your data | What is stored and where |
| 13 | Risks | Read this before you install |
| 14 | Troubleshooting | When something breaks |
| 15 | FAQ | Including what an MCP server is |
1. What you can ask it
- Post this, and put the link in a card rather than as bare text.
- Turn these notes into a thread. Show me the draft first, then stage part one so I can see it before anything is public.
- Which of my posts this month actually worked, ranked by engagement against views rather than raw likes?
- Read every reply I got today and tell me which ones deserve an answer.
- Publish these six screenshots as a carousel with alt text on each.
- Hide that reply, and everything nested under it.
- How much of today's posting quota have I used?
- Where are my followers, by country?
- Search for what people are saying about this launch, ranked by engagement.
- Restrict this post to the UK and Sweden.
The third one is the point. Threads reports views alongside likes, replies, reposts and quotes, so engagement can be measured against reach instead of against nothing. Ranked by raw likes, your best post is usually just your oldest.
2. Install
The long version, every step with what to do when one fails, is in INSTALL.md.
Node 22 or newer. Nothing else.
Authorise first, in a terminal:
export THREADS_APP_ID=... # from your Meta app
export THREADS_APP_SECRET=...
npx -y @thenavidm/threads-mcp-cli login
That stores a 60-day token at ~/.threads-mcp/tokens.json, and every client below picks it up with no credentials in its config at all. Section 3 covers where the app id and secret come from.
Claude Code
claude mcp add threads -- npx -y @thenavidm/threads-mcp-cli
A terminal
npm install -g @thenavidm/threads-mcp-cli
threads-cli
threads-cli install claude-code (or codex, claude-desktop, cursor, vscode, gemini) adds the server to a client in its own format; add --dry-run to see the change first.
That gives you two commands: threads-mcp is the server your AI tools launch,
and threads-cli is the one you type. Both are the same program.
Claude Desktop
The quickest route is the extension: download the
.mcpb from the
latest release and double-click it. No config file to edit. Leave its token
field empty and it picks up the refreshable one login wrote.
To wire it up by hand instead:
1. Open the config file.
In Claude Desktop, go to Settings, then Developer, then click Edit Config. That reveals claude_desktop_config.json in your file manager. Open it in any text editor.
If you would rather go straight there:
| System | Config file |
|---|---|
| macOS | ~/Library/Application Support/Claude/claude_desktop_config.json |
| Windows | %APPDATA%\Claude\claude_desktop_config.json |
| Linux | ~/.config/Claude/claude_desktop_config.json |
On macOS you can open it from a terminal with:
open -e ~/Library/Application\ Support/Claude/claude_desktop_config.json
2. Add the server.
If the file is empty or does not exist, paste this whole thing in:
{
"mcpServers": {
"threads": {
"command": "npx",
"args": ["-y", "@thenavidm/threads-mcp-cli"]
}
}
}
If you already have other servers, add only the "threads": { ... } part inside your existing "mcpServers", and put a comma after the entry before it. The file has to stay valid JSON. A single missing comma or a trailing one stops every server from loading, not just this one.
No credentials go in this file, because login already stored the token. If you would rather keep it here instead, add an env block with THREADS_ACCESS_TOKEN, and read section 10 first: a token in a config file cannot be refreshed by anything, so it dies on day 60.
3. Restart properly.
Quit Claude Desktop completely and reopen it. On macOS closing the window is not enough, use Cmd+Q. On Windows quit it from the system tray. Claude only reads that file at startup.
4. Check it worked.
Look for the tools icon in the message box and click it. You should see threads with its tools listed. Then ask it something from section 1.
If nothing appears, Claude Desktop's own log is the fastest way in:
| System | Log file |
|---|---|
| macOS | ~/Library/Logs/Claude/mcp-server-threads.log |
| Windows | %APPDATA%\Claude\logs\mcp-server-threads.log |
tail -n 50 ~/Library/Logs/Claude/mcp-server-threads.log
Two things account for most failures. Node is not installed, or not on the PATH that Claude Desktop sees, in which case use the full path to node as the command. Or the JSON is malformed, which you can check by pasting the file into any JSON validator.
Cursor
Create ~/.cursor/mcp.json for every project, or .cursor/mcp.json inside a single project. Use the same JSON as Claude Desktop. Then reload the window, or open Settings, MCP, and toggle the server.
Windsurf
~/.codeium/windsurf/mcp_config.json, same JSON, then reload.
VS Code
.vscode/mcp.json in a project, or run MCP: Add Server from the command palette.
Everything else
Zed, Cline, Continue and anything else that speaks MCP over stdio all work. They each keep their config somewhere different, but they all want the same things: the command, the args, and optionally the env.
Docker
The token store has to be mounted, or the container authorises into a filesystem that disappears:
docker build -t threads-mcp .
docker run --rm -i \
-v ~/.threads-mcp:/home/node/.threads-mcp \
threads-mcp
Self-hosting over HTTP
For a machine that is always on, which is also the most reliable way to keep a token alive:
THREADS_HTTP_PORT=8787 \
THREADS_HTTP_TOKEN=$(openssl rand -hex 32) \
threads-mcp --http
Binds 127.0.0.1 by default, and will not start on any other address without THREADS_HTTP_TOKEN. A Threads token can post as you, so put it behind a reverse proxy with TLS before you change THREADS_HTTP_HOST. GET /health returns the name, version and tool count without authentication.
Check it worked
npx -y @thenavidm/threads-mcp-cli doctor
It checks the network, then each token, then probes every capability separately: publishing, replies, insights, keyword search, profile discovery, geo-gating. Each one reports granted, missing, or missing with the exact scope to add.
3. Connect your account
Threads has no app passwords. Every credential is an OAuth token minted against a Meta app you own, which is a real setup step, so here it is in full. It takes about ten minutes once.
Create the app
[!TIP] One app covers Facebook, Instagram and Threads.
Use cases are ticked in a list, and you can tick several. If you plan to use more than one of these, do it now rather than making three apps and managing three sets of credentials.
Use case For Server Manage everything on your Page Facebook Pages facebook-mcp Manage messaging and content on Instagram instagram-mcp Access Threads API Threads this one Incompatible combinations grey out. If an option will not tick, it conflicts with something already selected.
-
Go to developers.facebook.com/apps and Create App.
-
Choose the Threads API use case.
-
In the app, open Threads API, then Settings. Copy the Threads App ID and Threads App Secret.
-
Under Redirect Callback URLs, add:
http://127.0.0.1:8788/callbackThat is the loopback address
loginlistens on. It never leaves your machine. If port 8788 is taken, uselogin --port=9000and add the matching URL instead. -
Under Roles, add yourself as a Threads Tester, then accept the invitation from your Threads profile at Settings, Website permissions, Invites.
Step 5 is the one people miss. Without it, every call comes back empty and nothing explains why.
Authorise
export THREADS_APP_ID=1234567890
export THREADS_APP_SECRET=abc123...
threads-mcp login
That opens the authorisation page, catches the redirect, exchanges the code, exchanges the short-lived token for a 60-day one, verifies it against your profile, and writes it to ~/.threads-mcp/tokens.json at mode 0600.
For the permissions that need App Review, once you have them:
threads-mcp login --all-scopes
If the browser cannot open, threads-mcp login --manual prints the URL and takes a pasted token instead.
The scopes
login requests these by default, and they work for you as a tester on your own app with no review at all:
| Scope | What it unlocks |
|---|---|
threads_basic |
Everything. Required for any call |
threads_content_publish |
Posting, threads, carousels, quotes, reposts |
threads_manage_replies |
Hiding replies, reply approvals |
threads_read_replies |
Reading replies and conversations |
threads_manage_insights |
Post and profile metrics, follower demographics |
threads_delete |
Deleting your own posts |
These three need App Review, and --all-scopes requests them:
| Scope | What it unlocks |
|---|---|
threads_keyword_search |
Searching posts other than your own |
threads_profile_discovery |
Looking up other public profiles |
threads_location_tagging |
Tagging posts with a location |
A missing scope usually shows up as an empty result rather than an error. threads_keyword_search is the worst of them: without it, Meta does not refuse a search, it quietly narrows it to your own posts. search_keyword notices when every result is yours and says so, and doctor probes for it directly.
Pasting a token instead
You can skip login and set THREADS_ACCESS_TOKEN to a long-lived token you already have. Everything works, with one consequence: the server has nowhere to write a refreshed token, so it cannot keep that one alive. See section 10.
Tokens from Meta's Graph API Explorer are short-lived and stop working in an hour. That is the single most common reason a Threads setup "randomly breaks".
4. What it costs to have connected
Both surfaces are the same program with the same 30 tools. The difference is when the model pays for them. Measured in Claude Code:
| Cost | MCP server | CLI |
|---|---|---|
| Every message, with every tool loaded | 11,400 tokens | nothing |
| Every message, Claude Code's default | 1,060 tokens | nothing |
| When Threads comes up | nothing more, or the tools it picks | 3,140 tokens for SKILL.md, once |
| 20 messages with Threads in 1, every tool loaded | 227,000 tokens | 3,140 tokens |
Claude Code's tool search is on by default: it sends only the tool names and the server instructions, and loads a tool's full definition when the model reaches for it. An app that loads every tool up front pays the first line on every message, whether Threads comes up or not. With the skill added, Claude Code also lists its one-line description, about 140 tokens.
Where the tokens go, with every tool loaded:
| Part of the tool list | Share |
|---|---|
| JSON Schema structure: types, required lists, nesting | 56% |
| Argument descriptions | 28% |
| Tool descriptions | 16% |
To spend less, turn the server off when you are not using it, which in Claude
Code is the /mcp panel. THREADS_READ_ONLY=1 takes the 12 write tools off the list, leaving 18.
Or install the CLI and add the server on the days it earns its place.
Measured on 2026-10-05 with Claude Code 2.1.286 on Claude Opus 5.5: one
short prompt with and without the server connected, once with
ENABLE_TOOL_SEARCH=false and once with the default, the difference read
from the API's own usage figures. SKILL.md was measured the same way, and the
shares were counted with OpenAI's o200k tokenizer. Other apps and models count
tokens a little differently.
Against 1.1.2, measured the same day: every tool loaded costs 11,356 tokens
instead of 12,745, tool search the same, and SKILL.md 68 more for the approval
rule, which and the full exit codes. In Codex 0.159.3 on gpt-6.1-sol, the same
task, "find the command that publishes a post staged earlier and the flags it
requires", read a median of 84,035 input tokens on 2.0.0 against 84,054 on 1.1.2
over the CLI (five runs each), and about 78,750 on both over MCP.
5. Tools
30 tools. Every one takes an optional account; every listing tool takes limit and cursor. Anywhere a post is named, it is the numeric id, which every read tool returns.
Accounts
| Tool | What it does |
|---|---|
list_accounts |
Every connected profile, which one acts by default, and days left on each token |
whoami |
Authenticate and return the live profile. Use this to confirm credentials |
get_publishing_limit |
How much of today's posting, reply and delete quota is spent |
refresh_token |
Extend this profile's token by another 60 days |
Posting
| Tool | Arguments |
|---|---|
create_post |
text, image_url, video_url, alt_text, link_attachment, topic_tag, reply_to_id, quote_post_id, reply_control, allowlisted_country_codes, enable_reply_approvals, confirm |
create_thread |
posts[], image_url, video_url, alt_text, link_attachment, topic_tag, reply_to_id, reply_control, confirm |
create_carousel |
items[], text, topic_tag, reply_control, confirm |
stage_post |
Everything create_post takes, minus confirm. Builds a container, publishes nothing |
publish_staged |
container_id, confirm |
get_container_status |
container_id |
quote_post |
text, quoted_post_id, confirm |
repost |
id, confirm |
delete_post |
id, confirm |
Replies
| Tool | Arguments |
|---|---|
reply_to |
id, text, image_url, video_url, alt_text, confirm |
get_replies |
id, reverse, limit, cursor |
get_conversation |
id, reverse, limit, cursor |
get_all_replies |
since_hours, limit, cursor |
hide_reply |
reply_id, hide |
get_pending_replies |
limit, cursor |
manage_pending_reply |
reply_id, action, confirm |
Threads exposes three different reply views and they are not interchangeable. get_replies is one level deep under one post. get_conversation is the whole tree under one of your posts. get_all_replies is every reply you have received across every post, which is the one you want when the question is "what needs answering".
Reading
| Tool | Arguments |
|---|---|
get_posts |
since_hours, since, until, limit, cursor |
get_post |
id |
since_hours reads a time window rather than a count: since_hours: 168 pages until it reaches a week back.
Insights
| Tool | Arguments |
|---|---|
get_post_insights |
id |
get_account_insights |
since, until, metrics[] |
get_follower_demographics |
breakdown (country, city, age, gender) |
get_top_posts |
sample, sort_by |
get_top_posts is the one that does not map to an endpoint. It fetches recent posts, pulls metrics for each, and ranks by engagement against views. That costs one request per post, so the sample is capped at 50 and the result says what it scored.
Profile insights only go back to 13 April 2024, and are unreliable before 1 June 2024. Earlier windows return nothing rather than an error.
threads-mcp-cli MCP Fragen
Welche Tools bietet threads-mcp-cli MCP?
Er stellt 30 Tools bereit, darunter list_accounts, whoami, get_publishing_limit, refresh_token, create_post.
Benötigt der Server Anmeldedaten in der Config-Datei?
Nein, Anmeldedaten werden nicht in der Config-Datei gespeichert, da der Login-Befehl ein refreshbares Token lokal speichert.
Wie lange ist das Threads Access Token gültig?
Das Token ist 60 Tage lang gültig und ist so konzipiert, dass es sich automatisch erneuert.
Gibt es Sicherheitsmaßnahmen beim Posten von Inhalten?
Ja, Aktionen wie Veröffentlichen, Antworten und Löschen erfordern eine Bestätigung, um versehentliche Posts zu verhindern.