We value your privacy
We use cookies to improve your experience, analyze site traffic, and show relevant content.
Strictly Necessary
Required for the website to function. Cannot be disabled.
Analytics
Help us understand how visitors use the site so we can improve it.
Marketing
Used to deliver relevant ads and track their effectiveness.
MetaPipe MCP · for Pipedrive
Your AI — the one you already pay for — handles your Pipedrive so you can fully focus on sales. Data entry, follow-ups, prioritisation, cleanup, quotes: it does the admin, you do the selling. No tutorials, no YouTube binge, no consultant on retainer.
🏆 Built by metawork.studio. Certified Pipedrive Authorized Partner — running Pipedrive for clients since 2016.
Built for two people
Most «AI for sales» tools assume you'll learn a new interface. MetaPipe assumes you're busy and need shit to be done. The AI is already on your screen, and the CRM should just do what you ask in plain convo.
For reps + founder-sellers
For sales directors + founders managing a team
Try saying this to your AI
Plain English. No field names, no IDs, no clicking around. Your AI handles the Pipedrive end, you keep selling.
"What's on my plate today?"
7 things. 3 stalling deals to nudge, 4 emails waiting on a reply, 2 quotes to send. Want me to draft the email replies?
"What stalling deals need my immediate attention?"
Acme ($48k, untouched 18 days), Nordwind ($22k, demo never followed up), Globex ($110k, decision-maker went quiet two weeks ago). Pulling the last conversation on each.
"Pull all the leads I need to reach out to today, then draft a personalised first-touch email for each one."
11 leads on today's list. Wrote a first-touch email for each — referenced the role, the company, and the angle from the source. Want to review before I queue them?
"Write a quote for Acme: 50 Pro seats, annual, 15% discount."
Done. €54,000 list, €45,900 after 15%. Filed under the Acme deal, attached to the existing thread with their procurement lead. Ready to send when you say go.
Everything it handles
There's no feature menu to learn — just things you can ask for, in plain words. Here's the range of what MetaPipe takes off your plate.
Full, hands-off configuration so you can onboard yourself — pipelines, stages, fields, automations, built to spec the moment you ask.
Pull exactly what you need out of Pipedrive and push exactly what you want back in — in plain language, no exports, no filters to build.
Hand it a messy spreadsheet and it maps the columns, cleans the values, and files every record where it belongs.
Keep the CRM tidy as you work — fix fields, fill gaps, reshape records — all on request, never by hand.
Ask for something and MetaPipe does the technical part the way we would do it for you. Right the first time.
A 30-second morning briefing from your real pipeline: what is stalling, who is waiting, what to send today.
Funnels, forecasts, win/loss, leaderboards — answered in chat, never built by hand in the CRM.
Pull lookalike leads from Apollo, craft sharp first-touch emails, and update the CRM — without ever opening it.
The progress you make in conversation lands back in Pipedrive as tasks, activities and projects automatically.
Roll out a sales campaign across your pipeline right where you are already typing.
A dedicated workspace that finds and merges duplicates fast. Fully GDPR compliant: your data never leaves your browser or reaches AI.
We constantly update the functionality and playbooks. Any update to Pipedrive, any change in their API, anything we add that simplifies and boosts productivity is included in free updates to your license.
The number that matters
~5h → 30min
per sales rep, per week, on CRM admin
The average rep loses about five hours a week updating Pipedrive. Logging activities, copying notes, chasing overlooked leads, figuring out the UI, cleaning up overdues, trying to find most important things. MetaPipe gives most of that time back, because the AI you already use can do it in seconds when asked.
For a team of five, that's roughly a full extra selling day per rep, every week. That's the headline. Everything else on this page is just how.
The strongest model you already pay for
Pipedrive's built-in AI assistant is one model, locked down, averaged to do things a certain way, no choice. MetaPipe is the other way round: it plugs into whichever AI you already use and pay for — Claude, ChatGPT, Gemini, Copilot, or any MCP-compatible client. When your AI upgrades to the next model, your CRM gets smarter the same day. No vendor lock-in to a model your sales team didn't pick.
Works with the AI you already use
Claude
ChatGPT
Gemini
Microsoft Copilot
GitHub CopilotUnder the hood, every common Pipedrive action is covered — from first-time setup to quotes to dedup to pipeline forecasts. You don't need to learn any of it. You just talk.
Setup
No code. No agency call. No consultant on retainer.
Pick a plan below and check out. 30 days free, cancel anytime.
In Claude, ChatGPT, or Gemini, add MetaPipe as a connector. We'll show you exactly where to click — under a minute.
Type connect to my pipedrive, authorize connection, you're done.
Detailed setup
Step-by-step for Claude, ChatGPT, and Gemini — plus Microsoft Copilot and GitHub Copilot.
Works in claude.ai (no install), Claude Desktop, and Cursor — same address, same login.
↓ Download Claude Desktop (optional)Claude.ai requires Pro, Team, or Max for custom connectors. On the free plan? Use Claude Code, Cursor, ChatGPT (Developer Mode), or Gemini CLI instead — same MetaPipe app, no subscription gate.
Pick a plan — see pricing below — and complete checkout.
Open claude.ai → click your profile picture → Settings → Connectors → Add custom connector.
Paste this address into the URL field: https://metapipe.metawork.studio/mcp/
Click Add. A sign-in screen opens — choose Google, Microsoft, or email, using the same address you used at checkout. MetaPipe recognises your license and connects automatically.
First time only: ask MetaPipe to connect your Pipedrive. Type connect my pipedrive and paste your Pipedrive API key when prompted.
Important: this requires the ChatGPT desktop app, not the website. The browser version of ChatGPT doesn’t support tool connections.
↓ Download ChatGPT Desktop (macOS / Windows)Pick a plan — see pricing below — and complete checkout.
Open the ChatGPT desktop app → Settings → scroll to the bottom → enable Developer Mode.
In Settings → Connectors → Add connector. Leave authentication set to OAuth.
Paste this address into the URL field: https://metapipe.metawork.studio/mcp/
Click Connect. A sign-in screen opens — choose Google, Microsoft, or email, using the same address you used at checkout. MetaPipe recognises your license and connects automatically.
First time only: type connect_pipedrive and paste your Pipedrive API key when prompted.
Gemini CLI is Google’s free AI assistant that runs in your computer’s terminal (command line). To use it, you need Node.js installed — it’s free and takes 2 minutes.
Pick a plan — see pricing below. Copy your license key from the confirmation email or your license account page.
Open your terminal (Mac: search «Terminal» in Spotlight · Windows: search «Command Prompt») and run: npm install -g @google/gemini-cli
Open the file ~/.gemini/settings.json in any text editor (create it if it doesn’t exist) and add the following — replacing YOUR_LICENSE_KEY with the key from step 1:
{
"mcpServers": {
"metapipe": {
"httpUrl": "https://metapipe.metawork.studio/mcp/",
"headers": {
"Authorization": "Bearer YOUR_LICENSE_KEY"
}
}
}
} Back in your terminal, type gemini and press Enter. MetaPipe’s tools will be available immediately.
First time only: type connect_pipedrive and paste your Pipedrive API key when prompted.
MetaPipe is a remote MCP server (Streamable HTTP + OAuth 2.1), so it plugs in with just its address and a browser sign-in. Add it in Microsoft Copilot Studio and use it inside your Microsoft 365 Copilot agents.
Heads-up: the consumer Copilot (Windows / copilot.microsoft.com) doesn’t support custom MCP servers — you connect through Copilot Studio.
Pick a plan — see pricing below — and complete checkout.
In Copilot Studio, add a tool, choose Model Context Protocol, and point it at https://metapipe.metawork.studio/mcp/. Sign in with the address you used at checkout, then add the agent to Microsoft 365 Copilot.
First time only: type connect_pipedrive and paste your Pipedrive API key when prompted.
MetaPipe is a remote MCP server (Streamable HTTP + OAuth 2.1), so it connects with just its address and a browser sign-in. Use it in GitHub Copilot agent mode inside VS Code or Visual Studio.
Pick a plan — see pricing below — and complete checkout.
In VS Code, open Copilot → Agent mode → Add MCP server and point it at https://metapipe.metawork.studio/mcp/. Sign in with the address you used at checkout.
First time only: type connect_pipedrive and paste your Pipedrive API key when prompted.
Why pay for this
Pipedrive’s own MCP is free, first-party, and genuinely fine for basic lookups — we’ll be the first to say so, and it has grown: leads are fully covered now. We ran it head-to-head against MetaPipe on the same account. It’s still a thin API passthrough: custom fields come back as raw hash keys (with no tool to resolve their names), records carry bare IDs instead of names and links, stages are read-only, no reporting, no duplicate cleanup, no guardrails, one account. MetaPipe is the operator layer that starts where the free one ends:
| Capability | MetaPipe | Pipedrive’s own MCP | Free community MCPs |
|---|---|---|---|
| Hosted + one-click sign-in (OAuth) — nothing to install or self-host | ✓ Yes | ✓ Yes | ✕ No self-host + paste API token |
| Read & write core CRM — deals, people, orgs, leads, activities, notes | ✓ Yes | ✓ Yes | – Partial a few objects |
| Covers the rest of your Pipedrive — products, projects, files, email, filters, goals, webhooks | ✓ Yes | ✕ No core records only | ✕ No |
| Understands your custom fields — names + option labels, not raw hash keys | ✓ Yes | ✕ No option labels yes, field names stay hashes | – Partial |
| Links every record — one click to open it in Pipedrive | ✓ Yes | ✕ No bare IDs, no links | ✕ No |
| Builds your CRM from scratch — pipelines, stages, custom fields, filters | ✓ Yes | ✕ No stages read-only, no pipeline tools | – Partial |
| Automations — Pipedrive exposes no API for them | – Partial build-ready spec, you click it | ✕ No | ✕ No |
| Reports on demand — funnels, forecasts, win/loss, leaderboards | ✓ Yes | ✕ No | ✕ No |
| Finds and merges duplicates safely | ✓ Yes | ✕ No | ✕ No |
| Protects your data — counts before/after, deactivate-not-delete, guided deletes | ✓ Yes | ✕ No updateDeal accepts is_deleted | ✕ No raw deletes, no checks |
| Our Pipedrive method baked in — does it the right way, not just what you literally typed | ✓ Yes | ✕ No | ✕ No |
| Email → CRM filing, daily briefing, summaries | ✓ Yes | ✕ No | ✕ No |
| Runs multiple accounts / clients from one seat | ✓ Yes | ✕ No | ✕ No |
| First-party, maintained by the CRM vendor | – Partial we maintain the adapter | ✓ Yes | ✕ No mostly solo side-projects |
| Price | from $29/seat/mo | free | free |
Credit where it’s due: Pipedrive’s own MCP is free, first-party, and genuinely fine for basic lookups and edits on core records. If that’s all you need, use it — we’ll tell you so. MetaPipe earns its price the moment the work gets real: understanding your custom fields, reporting, cleaning duplicates, building out the CRM, deleting safely, and running more than one account — none of which the free option does. And if you’re letting an AI work your CRM anyway, guardrails aren’t optional extras.
GDPR & your data
MetaPipe is built data-minimal by design. We don't keep a copy of your CRM, and the heavy lifting that touches real records runs in your own browser — so GDPR compliance is a side-effect of the architecture, not a checkbox.
Every request runs live against your Pipedrive and returns. We don't cache, copy, or warehouse your deals, contacts, or emails — there's no MetaPipe database holding your CRM.
The deduplication workspace analyses your records locally, on your machine. When you merge, only the record IDs travel to us to action the merge — never the record content — and those jobs auto-delete within 24 hours.
MetaPipe runs on servers in Germany. The only thing we persist is your connection credential — encrypted at rest — plus the email tied to your license.
The CRM data your AI reads to do its work goes only to the assistant you already chose — Claude, ChatGPT, Gemini — under your own provider agreement. MetaPipe adds no extra AI in the middle.
Pricing
A couple of hours with a Pipedrive specialist costs more than a year of MetaPipe — and MetaPipe runs your account every day, not just the hour you booked. 30-day free trial on both plans. No setup fees. Cancel anytime.
Single user
$€3929/mo
billed monthly billed annually — save 26%
Team
$€2922/seat/mo
billed monthly billed annually — save 24%
Changelog
Every license includes future additions — no tier upgrade, no extra invoice. Recent-first.
Updating or enriching thousands of records spends Pipedrive's daily API budget. That budget belongs to your whole company and every other integration you run shares it — use it up, and your sync apps and Make scenarios stop until it resets. Before a large run, MetaPipe now works out what it will cost, checks that against your plan and seats, and asks how you want it run: aggressive (fastest, may use most of the day's budget), moderate, or background (a process that runs on its own until the job is done, paced so nothing else notices). It also tells you when Pipedrive's own file importer could do the same job without touching the API budget at all.
A fix first: since MetaPipe moved to Pipedrive's newer API, a deal, person or organisation came back without its next and last activity, its note, file, activity and follower counts, and its won/lost deal counts — Pipedrive stopped sending those unless asked, and MetaPipe didn't ask. So "when is the next activity on this deal?" could get no answer, and a missing count looked like zero. They're back on every record and list. Alongside it, several smaller corrections: the recent-changes filter only offers kinds Pipedrive accepts, clearing a multi-select custom field works, file lists say when Pipedrive's 100-file cap applied, and a lead conversion that fails now says it failed. New: remove someone as a follower (and see follower history); create, edit, complete and delete project tasks; manage project custom fields; add and manage discounts on a deal; read products for many deals in one go; turn a deal back into a lead instead of deleting it; and search can now return product prices, related deals and leads, or match one field exactly. Every removal and the deal-to-lead move ask for confirmation first.
Three fixes found by re-reading Pipedrive's current API documents. Attaching or editing a product on a deal was still calling an address Pipedrive no longer publishes; it now uses the current one, so quoting a deal works again. The three subscription actions were calling an API Pipedrive removed in June 2025 — they now say so and point you at installments (or the Pipedrive screen) instead of failing with a raw error, and they no longer imply anything about whether a deal has a subscription. And when Pipedrive asks us to slow down, MetaPipe now waits and retries instead of passing the error straight through. That matters more than it sounds: ignoring those signals long enough gets your Pipedrive token blocked for every integration you own, and the block used to reach you as "reconnect your account", which is the one thing that cannot help. It now says what actually happened and that it clears on its own.
The Microsoft option sent you to an address that does not exist, so the sign-in could never finish — the URL was missing a segment Microsoft's current endpoints require. Self-hosted deployments can now also name the directory to sign in against (MICROSOFT_OAUTH_TENANT_ID); without it, an app registered for a single organisation is turned away. The allowed sign-in domain also became a list (METAPIPE_ALLOWED_DOMAINS), so a deployment can admit a customer's staff and the people supporting it without opening the door to every domain.
Since the 2026-08-27 API v2 migration every suggest_filing call failed before reaching Pipedrive with a Python scoping error, so nothing that relied on it — including automated inbox filing — could resolve a sender to their open projects and deals. Fixed, with a test that exercises the exact path and a guard against the pattern recurring in any module.
Listing a project's tasks showed done as empty on every task, because Pipedrive's v2 API names the field is_done on read (and done on write) and we were reading the write-side name. Tasks now carry done, milestone, start_date, priority and the time they were marked done. The phase column was also always empty: a task's phase is not on the task record at all in Pipedrive's API, it lives only in the project plan view, so MetaPipe now reads the plan alongside the task list and fills in each task's phase from it — which is also how the phase filter is applied. If the plan cannot be read, the list still comes back with phase unknown; only the phase filter itself refuses rather than quietly returning the whole project.
We compared every list and search tool against Pipedrive's published v2 API specification and closed the gaps. Activities can now be listed for a person, an organisation or a lead (not only a deal), fetched by a list of ids, narrowed to what changed since a point in time, and sorted by due date, id, creation or update time — and when you scope by a record, you get everyone's activities on it, not just your own. Search gained exact matching, a result limit, and record scoping: deals for a given person or organisation, or by open/won/lost status; people at an organisation; leads for a person or organisation. One correction came out of the same audit: the phase filter on project tasks was being ignored by Pipedrive, so asking for one phase quietly returned every task in the project — it is now applied on our side and returns only that phase.
Listing activities used to mean walking one owner's whole list and paging through it, which on a busy account is slow and often not what you wanted — your team keeps saved activity filters for exactly these views. Both pd_activity(action='list') and pd_report(kind='activity') now take a filter_id, the id of any saved activity filter (pd_filters(action='list', filter_type='activity') lists them). Pipedrive evaluates the filter on its side, so a view that was minutes of paging comes back in one call. When a filter is given the result is the view exactly as the Pipedrive UI shows it: MetaPipe no longer narrows it to the caller's own activities, and pending vs. completed is only applied when you ask for it — pass owner_id or done explicitly to narrow further. Nothing changes for calls without a filter.
MetaPipe's tool catalog was our biggest compatibility problem: some AI clients cap how many tools an agent can carry, and a 114-entry catalog costs every chat real context before the first call. So the surface was consolidated — nothing was dropped, related operations were folded into single tools that take an action: one tool for Pipedrive Projects, one for activities, one for reports, one for the mailbox, one for pipeline and stage admin, one for payment schedules, and so on. Every capability that existed still exists, all the safety machinery came along unchanged — confirm prompts still name the exact Pipedrive account a merge or deactivation would land in, deletes still don't exist, and the audit log still records every write. If your AI still has the old catalog cached, almost every old tool name keeps working: the call is forwarded to its new home and the result says what to call instead. The four exceptions (the three merge tools and pd_sources, whose arguments changed shape) return an error naming the exact new call rather than guessing at a destructive translation. Disconnecting and reconnecting the connector refreshes the catalog and makes the notes go away. Honest footnote: the token cost per chat drops by roughly a quarter to a third, not by the 61% the tool count suggests — the consolidated tools carry bigger descriptions — but the count is what fixes client tool caps, and 45 clears every cap we know of.
Self-hosted deployments store their state in PostgreSQL, and when that database was unreachable — a stopped server, a network hole — the service used to hang during startup forever: the health endpoint never answered, so a plain database outage looked like anything but one. Startup is now bounded. The service comes up degraded, /health answers ok:false with db "postgres:unreachable" and HTTP 503 — loudly distinguishable from healthy, so a deploy pipeline's health gate still refuses to promote a build that can't reach its database — and the logs name the actual connection error. Once the database answers again, the service finishes its deferred boot work and returns to healthy on its own, no restart needed. Every database connection attempt is also bounded now (METAPIPE_DB_CONNECT_TIMEOUT, default 10 seconds). Hosted deployments on SQLite are untouched.
Pipedrive took 69 of its v1 endpoints out of support on 2026-08-01 — everything MetaPipe used for reading and writing deals, contacts, organizations, products, activities, pipelines, stages and cross-record search. They still answer today, but Pipedrive can switch them off without notice, and their official docs already dropped them. Every affected call now uses Pipedrive's API v2, and results keep the same shape as before — including the linked names (which contact, which organization, which deal) that v2 stopped providing and MetaPipe now looks up itself. Two honest limitations came with v2: stage and pipeline positions can no longer be set through the API (Pipedrive removed that — reorder in Settings → Pipelines; new stages are appended in the order you create them), and re-activating a deactivated pipeline still relies on the last remaining v1 call, because v2 has no way to do it. Deactivating pipelines, the thing the data-safety rules actually require, works fully on v2.
A sweep of Pipedrive's recent API additions, adopted where they make MetaPipe better. Webhooks can now watch much more than records: deal line items (products and installments), Projects entities (projects, tasks, boards, phases), and — new — custom-field definition changes on deals, persons, organizations, products, leads and projects, so an integration can react when an admin reshapes the schema, not just when data moves. Webhooks can carry a name so you can tell them apart in Pipedrive's interface, and the webhook list now shows each hook's name and version (legacy 1.0 hooks deliver a different payload shape than the 2.0 default). Listing persons takes a deal filter — “who is on this deal?” is one call returning every linked person, participants included. Deal, person and organization lists fetched through the newer API come back with label names resolved, not just label ids. And notes now attach to Pipedrive project tasks: filter notes by task, or pin progress right where the work is tracked.
Pipedrive recently added outcomes to activities — Connected, Left voicemail, No answer on calls; Completed, Rescheduled, No show, Cancelled on meetings and other types your admin enables. MetaPipe now speaks them everywhere: log a held call in one step (create it done with its outcome), set or change the outcome on an existing activity by its plain label, and see outcomes in activity lists and in the activity report, which breaks completed work down by outcome — including how many activities were finished without one, which is your logging coverage gap. Pipedrive can require an outcome in its own interface before an activity is completed, but does not enforce that for connected tools, so MetaPipe reminds instead of failing: completing a call or meeting without saying how it went comes back with a nudge listing your account's outcome labels.
Asking MetaPipe to create a task, call or meeting and link the person it involves could fail outright — Pipedrive changed the activity API so the contact can only be set as the activity's primary participant, and a task that named a person was rejected. Since attaching the person is exactly what MetaPipe tells itself to do, the failure hit the normal path, not an exotic one. It now attaches the contact the way Pipedrive wants. Changing the contact on an existing activity works the same way and keeps everyone else already on it.
On a licence that covers several Pipedrive accounts, pd_use_account reported the switch, handed back an account_token to keep the chat on that account — and then the next write was refused for having no account selected. The token was the only thing tying the two calls together, and no tool advertised it as something you could send, so assistants dropped it. The switch now also holds for the rest of the connection on its own, and the token is a documented argument on every tool, so a chat stays on the account it picked even when the connection is shared with another chat or restarted underneath it.
A licence covers one Pipedrive company, and everyone on it connects to that same company individually, so MetaPipe acts as each person with exactly their own permissions. Two places got that wrong. Asking MetaPipe to connect counted company slots, saw the one slot in use, and answered that the licence was full — with removing a colleague's connection as the only way forward, which would just have moved the outage. And the error you hit mid-work pointed at your account page, where a team licence whose company is already connected offers nothing to press. Now both hand you a one-click link that connects you, personally, to the company the licence already uses: no extra slot, nothing taken from anyone, and no licence key travelling through a chat to get there. The account page also explains this on the Pipedrive card, and the Team members card can mint a connect link for any teammate who needs one.
If the one-click Pipedrive connection doesn't work for you, the fallback used to be typing a Personal API token into a conversation with your AI — which leaves a live CRM key in that conversation's history. Your account page now has a field for it instead: paste the token there, we verify it with Pipedrive, encrypt it, and never show it again. MetaPipe prefers a token stored this way over one pasted in chat, so if you already have one in a chat you can just add it on the page and the old one stops being used — nothing to disconnect. On a team licence you'll be asked to confirm one thing first: a token added this way is used by everyone on the licence, so your teammates reach Pipedrive as its owner rather than as themselves until you remove it.
Connecting Pipedrive by pasting a Personal API token means typing a live CRM key into a conversation, and it stays in that conversation's history — including anywhere the chat is exported, screenshotted or shared. We can keep credentials out of our own logs and out of tool results, but not out of your message. So connect_pipedrive now says so after the fact: the connection works and the token is fine, but if the account matters, treat the key as exposed, rotate it in Pipedrive and reconnect with the one-click install. The one-click link is now what MetaPipe offers first everywhere, with the reason attached instead of just a preference. Pasting a token still works and is not going away — some setups have no browser to run an install in.
MetaPipe can reach your Pipedrive two ways: a Personal API token you paste with connect_pipedrive, or the one-click Marketplace app. The rule is now stated plainly: a token you pasted is what gets used, and the app connection is the fallback for when you have not pasted one. Nothing changes for anyone who only ever used the one-click install. Worth knowing if you have both: reinstalling the app will not switch you back on its own, because your pasted token keeps winning — run disconnect_pipedrive() to drop it, or ask support, who can take the connection over from their side. whoami now reports which of the two is actually in play, so it is no longer guesswork.
Pipedrive fixes the permissions an app has at the moment you approve it. If the app later asks for more, an existing installation keeps the old set forever — nothing on our side can widen it, and reconnecting is the only repair. Until now that showed up as tools failing one at a time with a permissions error and no explanation. Now: if your connection is missing everything MetaPipe needs, the first call stops and hands you a one-click reinstall link instead of letting you discover it the hard way. If it is only missing some, nothing is blocked — whoami names the missing permissions and which tools they affect, so you can decide whether to reconnect. After you reinstall, the new permissions take effect within about a minute.
One chat can be pointed at several Pipedrive accounts, and each answer carries a small block of metadata telling your AI which one is active. Until now that block said nothing at all when no account had been picked — it left the field empty. But empty did not mean the call went nowhere: it went to the default account. So a chat that had been pointed at a client's Pipedrive and then lost that pointer (a server restart is enough) looked like it had no account selected, while it was quietly reading and writing somewhere else. The field is never empty now. If no account was picked, it says so and names the one being used instead, marked as the default so it can never be confused with an account you chose. Same for MetaHub, where no account selected means the call would fail rather than fall back — it now says that too, instead of leaving your AI to infer it.
pd_use_account lets you point one chat at a specific client's Pipedrive, and metawork_context lets you name that pointer so several chats can deliberately share it. Both work by handing your AI a signed token to carry on every following call — the token, not the connection itself, is what keeps the chat locked to that account, because the connection your AI client happens to be using can change without warning: a reconnect, a network hiccup, or us deploying an update all hand out a fresh one. Until now, naming a pointer with metawork_context (or pointing pd_use_account at one by name) skipped that token, so it relied on the connection staying the same after all. When it didn't, the chat fell back to your default Pipedrive silently — no error, just the wrong account. Every named pointer now gets a token too, so reconnects no longer reset it.
A Pipedrive address is not one value — it is street, house number, postal code, city, region and country stored as separate parts of one field, and those parts are what postal-code, city and country filters read, along with the map view and every regional report. Two things about them bite hard, and MetaPipe now knows both. Write an address as one line and Pipedrive stores the line without working out the parts: the record looks perfect while every filter that depends on it comes back short. Write only one part — just the postal code, say — and Pipedrive does not fill that part in and leave the rest alone; it replaces the entire address, so street, city and country are gone. MetaPipe now closes both. Changing an address reads the record first and merges your change into the parts already there, so correcting a postal code corrects the postal code and leaves the street alone — the fragment that would have blanked the rest never reaches Pipedrive. (Pass an empty part explicitly when you do want it cleared.) For anything past a handful of records it will steer you to the file importer with one column per part instead, which is the only path that fills the parts reliably — and to pause your webhooks and integrations first, because an import writes once per row and a sync running the other way can empty the field you are migrating from. This is a Pipedrive matter — HubSpot keeps those parts as ordinary properties and never had it.
MetaPipe asks you to confirm before anything irreversible — merging two people, cancelling a subscription, replacing a record's labels, updating deals in bulk. Until now those prompts named the record and stopped there: "pass confirm=True to merge person 33 into person 44" reads exactly the same whether it is about to happen in your own Pipedrive or in a client's, and one chat can be pointed at several accounts. Every confirmation now says which Pipedrive account it would act in, and hands your AI that account as a separate field so it can show it to you before you answer. The receipt after the write names the same account, matching what the audit log recorded. If the account cannot be worked out, the prompt says so loudly instead of quietly leaving it out.
`whoami` — the check your AI runs first to confirm it's connected to the right account — described both Pipedrive and HubSpot, whichever product you bought. On MetaPipe that meant every check also reported on HubSpot: a line saying no HubSpot is connected, and suggesting you install MetaHub from the HubSpot Marketplace — inside a Pipedrive product that has no HubSpot tools to use it with. Your AI reads that answer, so it could pass the suggestion on to you unasked. It now describes the CRM this product actually drives: your Pipedrive connection, which account this chat is pointed at, and who on your team has connected. Same for MetaHub in reverse. Nothing else about the check changed.
Every connection now reports which version of the tool set it is serving, and your AI can see it. That matters because of a quiet failure the protocol allows: your AI client caches the list of MetaPipe tools when it connects, and nothing ever tells it that list went out of date. When we rename or merge a tool, a client still holding the old list keeps offering the old name — and the call fails at the moment you need it, with an unhelpful "unknown tool". From now on: the version and a short fingerprint of the tool set ride along with every connection, `whoami` reports them, and a call made under a retired name either runs anyway (we forward it) or tells you exactly which tool replaced it and that reconnecting will refresh your list. Nothing about your CRM data changes; this is about MetaPipe never breaking silently underneath you.
Solo licenses cover one Pipedrive account at a time, and moving to another one was always allowed — but once you had connected an account the page offered no way to do it, so the only route was reinstalling MetaPipe from the marketplace. There's now a "Connect a different Pipedrive" button on your account page. Sign in to the company you want before you press it: it replaces the one you had, and MetaPipe stops reaching the old one.
Connecting a Solo license by pasting your license key into a config file no longer works — you sign in through your browser instead, the same way Claude and ChatGPT already connect. A license key identifies the license, not the person holding it: one Solo key pasted into five configs looks exactly like one person working from five machines. Signing in fixes that at the root. If you connect MetaPipe through Claude, ChatGPT, Microsoft Copilot or GitHub Copilot you were already signing in and there's nothing to do; Gemini CLI now signs in too. Team plans keep license-key access; they're seat-counted and tied to their CRM account already.
Setting up Gemini CLI used to mean pasting your license key into ~/.gemini/settings.json. It signs in through your browser now, like every other client: add the server, run /mcp auth metapipe, and approve it in the window that opens. Nothing secret sits in a config file on disk anymore.
Agency licenses no longer need a hand-copied Personal API token for each extra Pipedrive account. Ask your AI to connect and it hands back an authorization link — approve it in Pipedrive and that account joins your authorized set. Disconnecting works per account too, so removing one client's Pipedrive leaves the rest in place. Before it answers, each chat verifies which account it's actually pointed at, so a question about one client can't be answered out of another client's CRM.
For agencies and multi-practice operators: a license can now be authorized (per license, by us) for more than one Pipedrive account. Connect each additional account with its Personal API token, switch any chat between them with pd_use_account — by company id, domain, or name — and pd_list_accounts shows every authorized account with its connection status. Each chat carries its own account pin, so parallel chats never bleed into each other's CRM, and every account stays server-verified against the license.
Every entry in metapipe_audit_log now records WHICH Pipedrive account the write dispatched against — the thing you need once several accounts share one license or a team operator switches between clients. Filter with account= to see one account's history.
Deals join persons, organizations and leads as a fourth dedup entity — same review-and-merge flow, same audit trail. Recurring deals are handled: two deals that are members of the same series (monthly retainers, renewals) are recognized as siblings, not flagged as duplicates.
The paste-an-API-key mode of the dedup page is now gated behind a client invite link or Google/Microsoft/email sign-in, and every merge is attributed to the person who clicked it in the audit log. Same capability, no more anonymous merges.
Audit entries older than two years are pruned automatically — long enough to answer any «why did this change», short enough to be a sane retention policy.
Setting a discount on a deal's line item — via pd_add_product_to_deal or pd_update_deal_product — used to report success while silently dropping the discount, because Pipedrive retired the field these calls wrote to back in 2023 and the write just went nowhere. Discounts land for real now, as a percentage off the line or a flat monetary amount.
Last week's activity log didn't catch merges made through the dedup tool — they run on a separate path and slipped past it. They're logged now too, one line per merge naming the record that was merged away and the one that survived, so «why is this duplicate gone» has an answer there as well.
Every write your AI makes — updates, archives, merges, pipeline changes — now lands in an activity log you can ask about directly: «why did we change that» or «why is this deal missing» gets a real answer instead of a guess. Tell your AI why you're making a change and it notes the reason down too, so it survives past the conversation it happened in. Available on every plan.
When a lead qualifies, tell your AI to convert it and MetaPipe turns it into a deal through Pipedrive's own conversion — the lead's data, owner, and linked person and organization carry over the same way as converting in the Pipedrive UI. You get the new deal's link back right away; if Pipedrive is still processing a large conversion, MetaPipe hands you a conversion ID so you can check the result a moment later instead of guessing.
Listing deals, contacts, organizations, and products got three new handles: filter by when records were last updated («deals touched since Monday»), sort by creation or update time, and fetch a specific batch of up to 100 records by their IDs in one call. Follow-up questions like «which of these changed this week?» stop being full-CRM walks and come back in one round-trip.
The links MetaPipe gives you for email threads and messages now open the right place in Pipedrive's Sales Inbox instead of landing on a 'page not found'. Whether the email lives in your inbox, sent, drafts, or archive, the link points to the correct folder — and individual messages now come with a working link too, not just the thread. Ask your AI to find or file an email and the link it hands back just works.
Organization and person addresses can now be set piece by piece — country, postal code, city, street — instead of only as one full line. Ask your AI to 'set the country to Germany on Acme' and exactly that gets written, with the components stored properly so country and postal-code filters and reports pick them up. Full one-line addresses still work, custom address fields behave the same way, and a typo in a component name comes back as a clear correction instead of a cryptic API error.
When you merge duplicate leads in the dedup tool, MetaPipe now moves the losing lead's notes and activities onto the lead you keep, then removes the duplicate — nothing gets stranded. Pipedrive has no built-in lead merge, so MetaPipe does the reassignment for you. Persons and organizations already merged this way; leads now match.
Attaching a file to a deal, person, or organization no longer crawls on anything larger than a tiny document. Hand your AI a link to the file — a Google Drive or Dropbox 'anyone with the link' URL, or any public/signed link — and MetaPipe fetches it and attaches it directly, so the file never has to stream through the AI. Small files dropped straight into chat still work as before. Ask 'how should I upload files?' to see the fastest path for your setup.
Big imports now start with a clean-up pass. Drop your contacts, organizations, or deals sheet into the dedup tool and it tidies and de-duplicates the rows for you — matching exact first, then fuzzy, and ignoring legal suffixes and industry boilerplate when it compares — then hands back an import-ready CSV with your original columns intact: the winning record kept, absorbed duplicates removed. The method is built into the playbook too — ask your AI to plan an import and it steers large files through clean-then-import, counting rows in vs. out so nothing is silently lost.
Ask your AI to design a pipeline, model custom fields, plan an import, or run any task the Metawork method covers, and it now follows a documented playbook instead of improvising. pd_playbook() lists the topics (import · pipeline_design · custom_fields · filters · activities · data_safety · crm_choice · tools); pd_playbook(topic) returns the full method. Field-group hygiene is baked in — it checks existing field groups and asks where to file before creating anything.
New pd_create_pipeline, pd_update_pipeline, pd_create_stage, and pd_update_stage tools — stand up a new pipeline or reshape its stages without leaving the chat. Guard-railed by the method: a stage can't be deleted out from under its deals, and pipelines deactivate rather than delete.
pd_create_filter and pd_list_filters let your AI build and reuse Pipedrive saved filters, including multi-group OR logic and conditions on address components (postal code, city, country) that aren't first-class fields. A new filters topic in the playbook documents the address-subfield gotcha so conditions actually match instead of silently returning nothing.
pd_deals_summary rolls open / won / lost value, counts, and movement into a single answer, and the daily-briefing and forecast tools now assemble their data in one round-trip instead of many. «How's the pipeline looking?» returns in seconds.
Pipedrive API v2 search plus new reads and writes for followers, note comments, and field metadata. Destructive operations stay on the no-DELETE path — they route through pd_deletion_recipe, never raw deletes — so your AI can act on more of Pipedrive without falling back to the UI.
Listings and reports used to stop at a hidden 500 / 5000-row ceiling, so «all lost deals this year» could quietly under-count. They now exhaust every page — the totals you get back are the real ones.
pd_create_activity now refuses to create a loose task: every call, meeting, or to-do must anchor to a deal, lead, or project, plus the person or organization it involves. Stops activities from landing where no one will ever see them again.
The dedup queue is for long work — matching takes minutes, triage can take days, and the tab tends to stay open across breaks. Sessions are now a 5-day sliding window (any activity extends them) instead of a 15-minute hard cap. On top of that, the page transparently re-mints the session token if the server returns 401 — so a deploy restart or a long break no longer ends with "Session token expired" the moment you click merge.
When you connect by pasting your Pipedrive personal token directly into the dedup page (instead of going through pd_dedup), the SPA now mints a server-side session token automatically. Practical impact: merges no longer fail with "Missing session token", and the new settings UI (visible fields, match rules, audit log, export/import) shows up. Settings persist per-token, so reconnecting with the same token brings your configuration back.
When a cluster has more records (or more visible fields) than fit on screen, the comparison table now scrolls horizontally instead of squeezing columns. The field-label column on the left stays sticky as you scroll so you don't lose context.
Save a snapshot of your dedup configuration to JSON and roll back at any time — Export downloads the full license doc (team defaults + everyone's personal overrides), Import replaces it atomically with the same validation a normal save uses. The new Audit log tab records the last 50 changes (who, when, which scope, which fields) so it's easy to spot when a teammate changed something that affects everyone's queue. Import is recorded in the log too.
Each discriminator / identifier / soft signal can carry an optional condition that gates when it fires: minimum value length, non-empty, or a regex pattern (case-insensitive full match). Useful for rules like «treat customer_number as a discriminator only when it's at least 5 chars» or «promote to definite if both VAT IDs match DE.+». Backward compatible — existing rules without conditions keep working unchanged.
Team-license members can now keep their own visible-field choices and match rules separate from the team defaults without forking the whole config. The settings modal grows a Team / Personal toggle: edit team defaults in the Team scope (everyone sees the change), or build a personal layer in the Personal scope that overrides the team field-by-field per entity. A Reset to team default button cleans up overrides you no longer want. Solo licenses don't see the toggle — there's no team to override.
Every team works the dedup queue differently — some live by VAT number, some by customer ID, some only ever look at job title and city. The new ⚙ Fields & rules modal (top of the page, also reachable from any expanded cluster) lets each license decide which fields show in the comparison view per entity, and define three buckets of match rules: discriminators (values differ → not a duplicate, cluster hidden), identifiers (values match → promoted to Definite), and soft signals (contributes to Likely/Possible). Settings are shared across the whole Team license so the queue looks the same for everyone; last-edited-by stamped so you can see who changed what.
Migrated to Pipedrive's brand-new Projects v2 API (shipped 2026-05-21). pd_list_pd_projects gained server-side filters (deal_id, person_id, org_id, phase_id, status) so «projects on the Acme deal» is one round-trip instead of a full-mailbox walk. pd_project_for_deal collapses to a single call. pd_get_pd_project + pd_update_project switched to PATCH with nested custom_fields. New tools: pd_create_project, pd_archive_project, pd_search_projects (full-text), pd_list_archived_projects, pd_get_project_changelog, pd_get_project_permitted_users — plus boards (list/get/create/update), phases (list/get/create/update), and project templates (list/get). pd_list_pd_project_activities + pd_list_pd_project_plan intentionally stay on v1 — Pipedrive's v2 doesn't cover those reads yet. Delete-project is not exposed as a tool by design; use pd_deletion_recipe for destructive ops.
«Brief me on Acme», «what happened with the Smith deal», «catch me up on this lead» — pd_summarize gathers everything your AI needs to compose a coherent recap in a single tool call: entity context, activities (open + done), notes, emails, files, and (in broad scope) the history attached to the linked contact and organisation. Asks you up-front whether you want narrow (just the entity) or broad (linked records too), and recent (past 6 months) or all-history — the AI can't silently pick for you, so the recap matches what you actually wanted.
pd_list_persons, pd_get_person, pd_list_orgs, pd_get_org, pd_list_leads, pd_get_lead, pd_list_deals, pd_get_deal now include a resolved custom_fields list by default — labels and option values from /personFields, /organizationFields, /leadFields, /dealFields are joined onto each record so the 40-character hex hashes don't leak into your AI's view. /…Fields metadata is cached 5 minutes per tenant, so this is O(1) extra requests regardless of `limit`. Opt out per call with include_custom_fields=False if you only need standard fields.
The settings.json snippet on this page used `httpTransport` which Gemini CLI doesn't recognise; the canonical shape is a top-level `httpUrl` + `headers`. Anyone who copied the old snippet would have seen «MCP server failed to connect» — apologies. The corrected snippet is below in the Gemini CLI tab and an end-to-end step-by-step guide is published on the MetaPipe YouTrack as MPIPE-54.
Large dedup runs no longer require the browser tab to stay open. Clicking «Merge this group» or «Merge selected» now hands the work off to the server — only the IDs travel, never the record content. You can close your laptop, the lid will keep merging. The dedup tab shows live progress and reattaches automatically when you come back. Job files hold only the merge plan (entity type + survivor + loser IDs) and your already-resolved API token; they auto-delete 24 hours after the job finishes so credentials don't linger on disk. Works on both MetaPipe (Pipedrive) and MetaHub (HubSpot).
«Find emails from Acme last quarter that had attachments.» «Anything unread from David this week?» «Threads mentioning the renewal terms.» MetaPipe now does all of these. pd_list_email_threads gained since / until / from_email / to_email / subject_contains / has_attachments / unread_only filters, with smart pagination that early-terminates at the date boundary — so «last year» finishes in seconds instead of timing out. New pd_search_emails tool runs free-text search across subject + snippet by default, or full message bodies with parallel fetches when you opt in. Nothing is cached: every call walks Pipedrive live.
MetaPipe is now an authenticator-only app in the Pipedrive Marketplace. Install from your Pipedrive sidebar → OAuth → lands at your MetaPipe dashboard (if you have a license) or the pricing page with email pre-filled (if you don't). No marketplace bloat — install gives you the URL + Bearer token to drop into Claude / ChatGPT / Cursor; the actual tools run from our MCP, not Pipedrive's UI.
pd_update_project — Pipedrive's Projects API only accepts PUT for updates (PATCH returns 404), so updates now go through a dedicated tool that handles standard fields plus custom-field hashes, including the «_currency» companion key for monetary fields. pd_confirm_scope — verify the active Pipedrive account before destructive writes when you've switched accounts via pd_use_account; refuses with a clear warning if the active slot has silently reset to the env default. No-op for single-account callers, so no friction outside team mode.
Two refinements based on real-account testing: (1) same name or same domain at different addresses (different country, city, or street) is treated as a group / branch offices and skipped — those aren’t duplicates. (2) Shift-click on a cluster checkbox selects every row between it and the previously-clicked one, the way every list UI works.
Pipedrive doesn’t expose Sequences via REST. When you ask the AI to «add this lead to my onboarding sequence» or «build a drip campaign», this tool returns the two real options: (1) create + enroll manually in the Pipedrive UI, or (2) set up a Workflow Automation that auto-enrolls records tagged with a marker label / custom field — the AI then handles the tagging. No fake APIs, no guesswork.
Ask your AI to summarize a deal, lead, contact, organization, or project — it pulls activities, notes, emails, files (and for broad scope, the linked contact + org history) and composes a recap covering what happened, what was discussed, outstanding next steps, and accomplishments. The tool asks first: narrow vs broad (many teams attach activities to the contact, not the deal), and recent (6 months) vs all-history.
A hosted browser tool finds duplicate persons, organizations, and leads in your Pipedrive. Fuzzy name matching with Levenshtein gating, E.164 phone normalization, marketplace + social-host blocklists, Pipedrive-aware BCC handling. Data never leaves your browser — analysis runs locally, merges call Pipedrive directly. Decisions persist across re-pulls, so groups you skipped or flagged don’t reappear.
Added pd_pipeline_funnel, pd_pipeline_forecast, pd_pipeline_movement, pd_activity_report, pd_conversion_report, pd_won_deals_report, pd_lost_deals_report, pd_new_leads_report, pd_stale_deals, pd_user_leaderboard. Ask your AI for any common Pipedrive report by name.
Same codebase, two deployments. Pipedrive tools live at metapipe.metawork.studio, HubSpot at metahub.metawork.studio. Buy what you need, no bundle pressure. Internal split via METAPIPE_SCOPE.
Drop MetaPipe into ~/.gemini/settings.json with a Bearer token. Same tool surface as the Claude + ChatGPT paths.
OAuth flow in ChatGPT Settings → Connectors. Email match against your checkout account; no separate token.
pd_get_field_visibility / pd_set_field_visibility — read and control which Pipedrive v2 custom fields are exposed per object. Useful when an account has a 200+ field schema and you want the AI to focus on a working subset.
pd_get_pd_project, pd_list_pd_project_activities, pd_list_pd_project_plan, pd_list_pd_project_tasks, pd_project_for_deal. AI can now operate the new Projects module alongside deals.
MetaPipe split off the studio’s internal brain MCP into its own multi-tenant, scope-gated package. Portal-token licensing for retainer clients, dedicated subdomain, opinionated tool surface — designed to be sold, not just used internally.
pd_propose_deal_from_email, pd_file_email, pd_create_deal_from_email. Forward a thread and your AI drafts a deal with title, value, expected close — one confirmation away from the pipeline.
Initial launch — 80 core tools across deals, leads, contacts, organizations, activities, products, fields, pipelines, stages.
The AI you already pay for, doing the CRM work you'd rather not. 30 days free, then less than two hours of a specialist's time for the whole year, cancel any time.
Start 30-day free trialRunning an agency or consultancy? MetaPipe also handles multi-account dispatch, per-client credentials, and tenant-isolated team mode — the way we built it for our own client work. Contact us →