superagnt_

Agent toolkit for builders: one MCP server for database, memory, schedules and data tools.

View on superagnt.comTry this MCPSponsored

Hosted MCP Server

npx add-mcp 'https://mcp.superagnt.com/mcp'

Installs into Claude Code, Codex, Cursor and more

Documentation

One MCP server.Every tool your agent wishes it had.

Social data, lead enrichment, web scraping, a real Postgres database, memory, schedules, queues and an inbox. One URL, plugged into the agent you already run. An agent with no hands is just a chat window.

» one key across every source» credits charged only on a successful response» MCP or plain REST, same surface» your agent turns its own tools on

Copy the setup prompt.Your agent does the rest.

works with any agent+ 8 more

Install superagnt for me. It's the superagnt plugin from the github.com/superagnt/plugins marketplace, plus the superagnt MCP server (on Claude it's a connector I add in my browser; elsewhere I sign in through my browser). The setup guide for each kind of agent is at https://www.superagnt.com/install

or follow the detailed setup instructions

01// the data layer

The data your agent cannot get from the open web, one call away.

Profiles, posts, followers, ad libraries and keyword search across the big social platforms, plus any page on the web as clean text. Nine sources on one key, structured JSON whichever one you hit. No developer accounts to apply for, no scrapers to keep alive every time a layout changes.

[

Instagram01

Profiles, reels, hashtag media, locations and explore.

data_instagram_*

](https://superagnt.com/tools/instagram)

[

TikTok02

Creator profiles, video stats, keyword search across videos and accounts.

data_tiktok_*

](https://superagnt.com/tools/tiktok)

[

Facebook03

Pages, group posts, Marketplace listings, video and the public ad library.

data_facebook_*

](https://superagnt.com/tools/facebook)

[

LinkedIn05

Company and profile enrichment, people and job search, posts and engagement.

data_linkedin_*

](https://superagnt.com/tools/linkedin)

[

Reddit06

Subreddit feeds, full comment threads, user history and search.

data_reddit_*

](https://superagnt.com/tools/reddit)

[

X (Twitter)07

Posts, profiles, followers, replies, search, lists and trends.

data_x_*

](https://superagnt.com/tools/x)

[

Web scraping08

Any page as clean markdown or schema-shaped JSON. Search and site maps too.

data_web_*

](https://superagnt.com/tools/web)

[

Email finder and enrichment09

Find and verify an email, find a mobile, enrich people and companies.

data_agnt_*

](https://superagnt.com/tools/leads)

» find anyone, verify anyoneseveral providers tried per call, one answer back

people

»search by role, company and city
»find a work email from a name and a company
»verify it is deliverable before you send
»find a mobile number when the email gets ignored
»turn a profile url into a full record
»one person, or a whole list in one pass

companies

»search by industry, headcount and geography
»find the companies that look like your best ones
»turn a domain into size, industry and location
»pull every public address on that domain
»watch hiring, funding and tech signals
»one domain, or a whole list in one pass

Priced per call, in credits, on the same key as everything else.

» the upstream bills are ours

Get started→

02// database

Give your agent a database.

A folder of markdown gets an agent surprisingly far, until the third time it re-reads everything to answer one question. Everything yours collects lands in an actual database it can create tables in, migrate, and query with SQL. Not a key-value toy: the memory layer real work sits on.

»tables, SQL, migrations and CSV loads: tools, not a dashboard
»every scrape, enrichment and job result lands somewhere queryable
»10 GB to start, isolated per workspace

agnt_db_* → on by default, the moment you connect

tool call

data_instagram_* posts · profiles

tool call

data_agnt_people_* enrich · find

tool call

data_web_* scrape · crawl

workspace postgres

agnt_db_insert

recordvia#

@lena.buildsinstagram001
sarah chenpeople002
northloop.comweb003
@marco.runsinstagram004
iris waltonpeople005

leads5 rows

» select count(*) from leads;5

inserting·3 sources wiredrows landed5

03// canvas

Live dashboards, built by your agent.

One tool call pins a query as a live widget, and the widgets add up to an operation. Below: a sales agent's own board, the funnel it is working, the pipeline it moved overnight, and the two decisions it parked for its founder. Shared with a link; nobody asks for an export again.

canvas // sales-agnt — pipeline ops

live · shared

pipeline value

$184.2k+$31k wk

demos booked · wk

replies · wk

new qualified · wk

New leads3

Contacted3

Replied4

Demo booked3

Closed won3

1,24041296229

33%23%23%41%

sentopenedreplieddemoclosed

montuewedthufri

$184.2k

Jul 6Sep 28

02:10pulled 212 new signups · enriched + icp scored212 rows

03:15sent 38 follow-ups in your voiceemail · li

05:02acme.com replied "send pricing" · logged + queuedcrm

05:40moved meridian labs to demo booked · thu 3:00pm$18k

06:31booked northloop.coffee demo · wed 11:00am$4.8k

06:58drafted meridian contract + flagged basalt pricing · parked for youapproval

agnt_canvas_*every widget above is the real component the product renders

04// imessage + slack

Talk to your agent in iMessage or Slack.

Bind your agent to a phone number and a Slack channel, then work with it like a person: text it from anywhere, get the answer in the same thread, and when it needs a yes, that yes is one reply.

»bind a phone number or a Slack channel, nothing to install
»answers land in the same thread, on your phone or in the channel
»approvals arrive as a message. one reply releases the send

agnt_bindings_*same agent, same memory, both surfaces

Claude Code

iMessage

05// email + inbox

Better tools for agentic inbox management.

Connect Gmail or Outlook once and the whole mailbox becomes tools: search, read, draft, reply, send, label, file. Triage and escalation come built in, and your agent only opens the messages that matter.

»triage and escalation built in: what needs you gets escalated, the rest gets handled
»it reads what matters and skips the rest, so a 200-email day stops costing tokens
»Gmail and Outlook connect once: search, read, draft, reply, send, label, file

connection_gmail_* · connection_outlook_*family: email · off until you turn it on

// your inbox

0 / 6 triaged

drafted

filed

surfaced

» inbox.triage --live08:01 AM

──triaging inbox──

» 200 in · 6 opened · 2 for youthe rest never entered context

06// knowledge + memory

What one agent learns, every agent knows.

Your agents share one memory. What the research agent learns on Monday, the outreach agent uses on Tuesday, because both write to the same graph of people, companies and projects. The usual alternative is one memory silo per platform: facts you cannot read, cannot move, and have to teach again from scratch the day you switch tools.

»any harness, any agent, one graph and one set of facts
»people, companies and your business in one place you own and can export
»ingest a document once; every agent retrieves it by meaning

memory on by default; knowledge switches on from the dashboard

peoplecompaniesprojectstopics

memories: 0 and counting

07// data jobs

Hand it forty thousand rows. It finishes without you.

A data job runs outside the chat. Your agent writes the recipe once (reformat this row, verify this email, enrich this signup) and the pipeline runs it as code over the whole list. Nothing to babysit: no context window to blow, no session to hold open, and items that fail get retried. Push those same forty thousand rows through chat instead and you pay a model to think forty thousand times.

»it runs on the pipeline, not in chat: no context window, no open session
»an item that fails is retried, and the job resumes where it stopped
»2,000 email checks cost 2,000 lookups, not 2,000 model turns

reformat 40,000 rows overnightverify 2,000 emails before the 9am sendcheck 9,000 careers pages every morning

agnt_data_job_*priced per item, not per token

your agent

agnt_data_job_submit callback received

reformat-rows

done

40,000/ 40,000

✓08,412reformatted
↻19,077retried, ok
✓33,190reformatted

table

leads40,000 rows

file

rows.csvwritten

webhook

your agentjob donewebhook back to your agent

job complete·3 outputs wiredmodel calls per item0

08// webhooks

Something happens out there. Your agent wakes up.

Your agent gets inbound webhook URLs. Point Stripe, GitHub, a form, or anything that can POST at them, and the moment an event lands a session starts with the payload in hand. No polling, no glue code, no you.

»a Stripe event lands, a session starts, the work happens
»payloads are verified per endpoint and handed straight to the session
»outbound too: your agent can send webhooks to anything with a URL

agnt_webhooks_* → every firing is a fresh session with a receipt you can read

inbound webhooks

3 endpoints · no session needed

stripe-events payment_intent.succeeded → payload verified · session started
github-push push · main → release notes drafted
signup-form form.submitted → lead enriched + queued

» fires whether you're around or notstripe · github · forms

09// custom mcp servers

Every agent gets its own custom toolkit.

Mint one MCP server per agent and hand it only the tools its job needs. The researcher gets the data sources. The outreach agent gets the inbox and the database. Nobody gets what it does not need. The alternative is one over-powered agent with every tool switched on, and no sentence in a system prompt stops that agent from using them.

research agent

digs through X, the web and your knowledge base

mcp•research-bot

» serves

data_x_*data_web_*agnt_knowledge_*

not served to this agent

connection_gmail_*agnt_db_*agnt_webhooks_*agnt_schedules_*agnt_queues_*

outreach agent

finds people, sends the email, logs the touch

mcp•outreach-bot

» serves

data_agnt_people_*connection_gmail_*agnt_db_*

not served to this agent

data_web_*agnt_webhooks_*agnt_schedules_*agnt_canvas_*agnt_files_*

reporting agent

reads the numbers, renders the recap

mcp•reporting-bot

» serves

agnt_db_* (read only)agnt_canvas_*

not served to this agent

connection_gmail_*data_x_*agnt_webhooks_*agnt_queues_*db writes

»only the tools the job needs· 01/03

»the researcher gets data and knowledge; it cannot send an email
»the outreach agent gets the inbox and the db; it cannot scrape the web
»revoke one agent's server at 2pm and the other two keep working managed from the dashboard, or by the agent that owns them

10// agent runtime

When one agent isn't enough, deploy a fleet.

The toolkit's ceiling: the agent you are talking to can build agents of its own. Deployed agents run on our managed runtime with their own prompts, tools, schedules and channels, and everything they do is observable, down to the individual tool call.

»your harness creates the agent; ours keeps it alive
»schedules, channels and a knowledge base wire in at deploy time
»sessions, tool calls and spend are all queryable after the fact

agnt_observability_* → families: lifecycle · bindings, off until you turn them on

pipeline // pipeline live enriching lead → company + 6 signals
inbox // triage live drafting reply → inbound from acme.co
research // signals live gathering intel → 12 sources
revops // revops live flagging for approval → refund $480
scheduler // scheduler live scheduling meeting → tue 14:30

11// what it costs

Every capability on every plan. You pick how much.

Same tools, same data sources, same database on all three. What changes is how much of it you get and how fast you can call it. Hit a cap and your agent offers the next plan. Nothing is metered past it.

[ free trial ]

No card. Verify your email and you get $3 of data credits and Growth limits. The 7-day clock starts at your 25th billable data call, your first use of the workspace database, or when the credits run out, whichever comes first. After that everything pauses, nothing is deleted, and your data is kept for 30 days.

Starter

Every capability, sized for your first agents.

$49/mo

$20 data credits included / mo

ai credits top up separately

1 workspace · 3 deployed agents
10 schedules · 3 queues
Database: 10 tables, 25,000 rows
1 inbox · 1 messaging line
60 requests / min
Every tool, data source and connection

Start free, no card

Growth

every limit	Starter	Growth	Scale
Price / mo	$49	$149	$399
Data credits included / mo	$20	$90	$300
Workspaces	1	3	10
Deployed agents	3	15	50
Schedules	10	50	250
Queues	3	15	50
Webhook events / mo	25,000	150,000	750,000
Data job pipelines	3	15	50
Data job items / mo	10,000	75,000	400,000
Knowledge collections	3	15	50
Knowledge documents	500	5,000	25,000
Knowledge storage	2 GB	20 GB	100 GB
Database tables	10	50	250
Database rows, all tables	25,000	250,000	2,500,000
Database storage	250 MB	1 GB	10 GB
Inboxes	1	3	10
Messaging lines	1	3	10
Requests / min	60	300	600

A credit is a cent.

One balance per organization, split into a data wallet for API calls and an AI wallet for the model tokens a deployed agent burns. Your plan fills the data wallet every month. The split is there so a runaway agent can't drain the wallet your lookups run on.

1 credit = $0.01

You pay per call, and only when it works.

Every billable response carries what it cost and both remaining balances in its meta block, so the agent can read its own spend mid-run. A call that fails is not charged, and a call the balance can't cover returns no data and costs nothing.

meta.costCents

Hit a limit and your agent offers the next plan.

Every cap is hard, including the data credits a plan includes. When a call would cross one, your agent gets an upgrade offer naming the next plan, what it unlocks and a link for you to confirm. Nothing is metered past the cap, and nothing you built is deleted.

requires_upgrade

AI is not included.

Plans cover the platform and the data. The agent you already run brings its own model. Agents you deploy here burn AI credits you top up in the dashboard, and that goes for the trial too. Building one with the Agnt Wizard costs no AI credits.

ai wallet

Per-source pricing is published.

Each source prices its own calls: a web fetch is a fraction of a cent, an enrichment that walks several providers costs more. The number for a given source lives on that source's page.

/tools/<source>

see per-source pricing → » balances are organization wide, keys are per workspace

12// read the rest

// documentation

Docs

Authentication, error shapes, pagination, rate limits, the REST surface and the MCP tool reference.

open→

// catalog

Every tool, one page each

A page per data source and per tool family: what is inside, the real tool names, and a worked call.

open→

// endpoint

One URL, every workspace

The same address for everyone. Your key decides which workspace it resolves to and what it can reach.

https://mcp.superagnt.com/mcp

» want it run for you?

Meet superagnt_

Same toolkit, with an operator already sitting behind it. You describe the job in plain language. It builds the agent, wires the tools, runs on a schedule and reports back.

Meet superagnt_→

13// questions

Answered before you paste a key.

all answered

·10 questions

01How does auth work?

OAuth first. The endpoint is its own authorization server, so almost every client (Claude Code, the Claude app, ChatGPT, Codex, Cursor, VS Code, Gemini, Windsurf, Zed, Grok, OpenClaw) runs the browser consent flow on first connect and gets its own credential. No token is ever pasted or revealed, and the consent screen doubles as signup. A workspace bearer token from the dashboard is the fallback for clients that still need one. The block at the top of this page prints the exact snippet for whichever one you run.

02Is there a free plan?

No. There's a free trial instead, and it doesn't ask for a card: verify your email and you get $3 of data credits and Growth limits. After that it's Starter, Growth or Scale, from $49/mo. Every plan has every capability. They differ only in how much you get.

03When does the trial start?

When you actually use it. The 7-day clock starts at your 25th billable data call, the first time you use your workspace database, or when the $3 runs out, whichever comes first. Connecting a client and letting your agent look around the tools don't count. If the credits run out early, only data calls stop and everything else keeps working until day 7. Then everything pauses, nothing is deleted, and your data is kept for 30 days so you can pick a plan and carry on.

04What counts as a credit?

A credit is a cent. Data credits pay for per-call data spend: data API calls and lead enrichment. Your plan includes a batch of them every month. AI credits pay for the model tokens and runtime a deployed agent burns, and you top those up in the dashboard. They are separate wallets on purpose, so an agent that runs long on tokens cannot empty the balance your lookups draw on. Credits come off on success, and every billable response reports what the call took and what is left.

05What happens when I hit a limit?

Your agent tells you. Every cap is hard, including the data credits your plan includes. When a call would cross one, the agent gets an upgrade offer naming the next plan, what it unlocks and a link for you to confirm. Nothing is metered past the cap, so there is no overage bill. Move to a smaller plan and anything over its caps pauses. Nothing gets deleted.

06Is AI included?

No. Plans cover the platform and the data. The agent you already run (Claude Code, Codex, Cursor) brings its own model, so that costs you nothing extra here. Agents you deploy on superagnt_ burn AI credits you top up in the dashboard, and that goes for the trial too. Building one with the Agnt Wizard costs no AI credits.

07Are there rate limits?

Yes. Each plan has a requests-per-minute ceiling (60, 300 or 600), shared by every key and every MCP connection in your organization. Over REST, going over returns a 429 with a retryAfter telling you how long to wait. Over MCP, your agent gets the same wait time plus an offer for the next plan. A bigger plan buys throughput as well as quantity.

08Is my data isolated?

Yes. Keys, database, files and memory are all workspace scoped. The Postgres database is provisioned per workspace with no connection string handed out, and memory never crosses a workspace boundary. Your data and your agent's memory are yours, and they are not used to train foundation models.

09Can my agent enable more tools itself?

Yes, and that is the intended path. A fresh connection starts on a small default surface so the tool list stays readable. From there agnt_tools_search finds what else exists and agnt_tools_enable switches a family on, so the agent widens its own toolset mid task instead of waiting for you to tick a box. agnt_tools_disable takes it back off.

10Do I need to be a developer?

No. For most clients the wiring is one line you paste into a terminal or a config file, and for the agent clients there is a prompt you can hand to the agent so it does the wiring for you. After that you talk to your agent, not to us. If you would rather not run MCP at all, every tool is also a REST endpoint you can call with the same key.

Get started→ » then paste the snippet from the top of this page