Paperclip
Paperclip is the open-source control plane for managing a team of AI agents — org chart, goals, budgets, and approvals in one app.
Paperclip is the most thought-through take on agent orchestration we've tested: output, permissions, and budget are product surfaces here, not afterthoughts, and Connectors plus per-user secrets push the governance model past what script piles or single-session runners manage. If you're technical enough to self-host and you're juggling more than two agents, the setup time pays for itself. If you need something live in ten minutes on someone else's servers, keep waiting on the hosted version.
Verified 5d ago · liveness 74/100 · cite: rightaichoice.com/tools/paperclip
- Technical builders running a multi-agent company across dev, content, social, marketing, QA, research, and outreach
- Small teams that need shared visibility, a Work Timeline, and per-user secrets on top of self-hosted agents
- Developers who want budget caps and approval gates before agents spend money or ship work
- Anyone replacing a pile of separate agent scripts and automations with one orchestration UI
- Single-purpose chatbots or one-agent automations that don't need org structure
- Non-technical operators who won't touch a CLI or self-host infrastructure
- Teams that need a fully managed cloud product today — the hosted version is still a waitlist
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Paperclip if you want a one-agent automation or a fully managed cloud product you never have to operate — it's CLI-driven, self-hosted, and the governance features only pay off if you work the approval queue.
You pay the infrastructure bill, not the vendor: database, hosting (Docker, Fly.io, or Modal), and secrets storage all land on your account.
Paperclip is open source under the MIT License, with a self-hosted local install available now and a hosted version on a waitlist. That puts the entry cost at your own infrastructure plus model-provider usage rather than a per-seat subscription — cheaper than managed orchestration services if you have the engineering time, more expensive in hours if you don't.
In short
Paperclip — Paperclip is the open-source control plane for managing a team of AI agents — org chart, goals, budgets, and approvals in one app. Best for Technical builders running a multi-agent company across dev, content, social, marketing, QA, research, and outreach, Small teams that need shared visibility, a Work Timeline, and per-user secrets on top of self-hosted agents, Developers who want budget caps and approval gates before agents spend money or ship work. Free to use.
What's new in Paperclip
Checked 4 days agoAcross the latest 10 updates: 2 feature updates, 3 changelog entries and 5 community discussions.
Guide: Running AI Agents Overnight
Paperclip guide covers what agents can finish unattended, what should wait for a person, which layer enforces limits, and morning checks.
Paperclip v2026.1012.0 released
Paperclip v2026.1012.0 is live; the release is flagged breaking. Details in the changelog upgrade guide.
Guide: Self-Hosted AI Agent Stacks
Paperclip compares MIT-licensed agent stacks, covering what you run, where model calls still go, and what your team maintains.
Guide: Agent Harness vs. Framework vs. Runtime vs. Control Plane
Paperclip guide maps which layer blocks a risky command, resumes a broken session, and tracks task ownership across Claude Code and Codex.
Paperclip v2026.1005.0 released
Paperclip v2026.1005.0 shipped, flagged as a breaking release. Upgrade notes in the changelog.
Guide: 7 Platforms for Managing AI Agent Teams
Paperclip roundup of seven tools for organizing AI agents, managing spend, and keeping decisions with the operator.
Connectors: Give your agents the access they need
Paperclip Connectors let you set which apps, accounts, and actions each agent can access.
What Is an AI Agent Control Plane?
Paperclip guide explains the five layers an agent control plane governs and how to evaluate ownership, approvals, budgets, and evidence.
Paperclip v2026.831.0 released
Paperclip v2026.831.0 shipped as a breaking release. Changelog has the upgrade details.
Your task page is now a conversation (v2026.824.0)
Paperclip v2026.824.0 turns the task page into a conversation and adds in-product Claude and Codex sign-in plus Tailscale HTTPS previews.
What people actually say about Paperclip — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
24 mentions across 6 sources (Hacker News, Product Hunt, App Store, Stack Overflow, GitHub, Lemmy), 48 more we could not attribute · researched Sep 24, 2026.
Weighted by the 72 posts each of 6 sources contributed.
- +Org-chart mental model maps cleanly to how teams already think about delegation and ownership
- +Per-agent budget tracking and approval workflows give real governance, not marketing theater
- +Interactive `npx paperclipai onboard` handles database and auth setup automatically
- +Model-agnostic across Claude, OpenAI, Gemini, and Codex avoids single-vendor lock-in
- +Interface quality repeatedly compared to Linear, which matters for daily operator use
- −Disclosed RCE via malicious agent imports makes self-hosting a genuine security risk
- −Single malformed comment ID silently blocks all wake dispatches for an issue
- −5,597 open GitHub issues signals maintenance is losing ground to feature velocity
- −Self-hosted only means no managed cloud option for non-technical operators
- −Name collision with Paperclips game and Rails gem makes finding honest reviews hard
- • LLM API spend across many agents can compound fast with no built-in cap beyond per-agent budgets
- • Self-hosting means you pay for compute, database, and secret-management infrastructure yourself
- • Security patching and DevOps time is an unlisted cost that scales with your deployment footprint
Viability Score
How well maintained and how widely used is Paperclip? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Org chart for managing multiple AI agents as a team
- Hire agents into roles from pre-built templates including Coder and Marketer
- Assign goals and tasks across agent roles
- Per-agent and per-team budget tracking and spend caps
- Approval workflows before agents execute work
- Decision requests: agents ask permission before acting
- Per-user secrets with API-based agent secret requests, logged for audit
- Connectors: scoped app, account, and action access per agent (September 15, 2026)
- Shared Work Timeline and unified Activity page across tasks
- Task pages that work as conversations
- Artifacts page indexing every agent work product with playback
- Bundled wireframe skill: one-line prompt to multi-screen SVG flow with published HTML viewer
- In-product Claude and Codex sign-in
- Tailscale HTTPS previews
- Task watchdogs for long-running agent work
About Paperclip
Paperclip is the app people use to manage AI agents for work. Rather than prompt a single chatbot, you build a company: hire agents into roles from pre-built templates (Coder, Marketer, and others), assign goals and tasks, and cap spend with per-agent and per-team budgets. The mental model is deliberately organizational — reviewers keep describing it as the company, where a single agent is only the employee. Technical builders running dev, content, social, marketing, QA, research, and outreach agents side by side are the core audience. It runs self-hosted and model-agnostic, and the interactive CLI onboard (`npx paperclipai onboard`) wires up database and auth for you; a hosted version sits behind a waitlist while the local build ships now. Governance is where Paperclip separates itself from script piles. Decision requests make agents ask permission before acting on specific calls, a shared Work Timeline and unified Activity page track coordination across tasks, and per-user secrets let agents request approved credentials on demand through the API with every request logged. Connectors, launched September 15, 2026, let admins pick exactly which apps, accounts, and actions each agent can reach instead of granting blanket access. Agent output is a first-class object too: the Artifacts page indexes every work product across issues and runs, grouped by task stack with playback, and a bundled wireframe skill turns a one-line prompt into a multi-screen SVG wireframe flow with a published HTML viewer. The release pace is brisk and occasionally sharp-edged. Task pages became conversations, in-product Claude and Codex sign-in landed, Tailscale HTTPS previews arrived, and the current build is v2026.1012.0 (October 12, 2026). Before that came v2026.1001.0 on October 1, 2026, flagged breaking, following the equally breaking v2026.916.0 and its v2026.916.1 patch. Budget an hour for changelog reading before you upgrade. Positioning-wise, Paperclip is broader than a
Behind the Verdict
We'd reach for Paperclip the moment a solo builder stops being able to hold every agent's state in their head. The org chart, shared Work Timeline, and unified Activity page are the parts that actually change behavior — they make agent work legible to a second person, which is the point where most DIY setups fall apart. Pick it when you're running multiple agents across different functions and you care who approved what. Decision requests and the credential-request flow (per-user secrets, logged on every request) mean an agent can't quietly spend money or touch a production account. Connectors tighten that further: per-app, per-account, per-action scoping instead of one broad token for everything. The Artifacts page is the sleeper feature. Indexing every work product by task stack, with playback, turns 'what did the agents actually produce this week' into a two-click question. The wireframe skill — one prompt to a multi-screen SVG flow with an HTML viewer — is the clearest example of Paperclip treating output as a deliverable rather than a chat transcript. Where it bites: the none-of-your-business parts. Releases are frequent and several recent ones were flagged breaking (v2026.916.0, v2026.1001.0), so pin a version and read the changelog before you pull. Self-hosting is a real commitment. And the governance layer only earns its keep if a human actually works the approval queue — wire it up and walk away and you've built an expensive audit log. Compared to assembling LangChain or CrewAI scripts plus a task board, Paperclip ships the org structure, budget caps, secrets handling, and artifact index already built. Compared to a single-session coding runner, it's a different category: that's one worker, this is the company around it. We'd pick Paperclip over a
Researching Paperclip? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Paperclip actually fits — and what changes day-one when you adopt it.
Run `npx paperclipai onboard` to stand up the instance and database, then build an org chart with Coder, Marketer, and QA agents, assign each a monthly budget, and connect GitHub for code and Ramp for spend.
Outcome: Dev, content, and QA work run under one dashboard with per-agent budget caps and an Activity log, instead of five separate scripts with no shared visibility.
Set up shared Work Timeline and per-user secrets so the support agent requests a scoped credential on demand, then review decision requests and approvals each morning in the Activity page.
Outcome: Agents can reach Slack and the CRM without a shared master credential, and every secret request is logged for review.
Give the bundled wireframe skill a one-line prompt describing the flow, then review the generated multi-screen SVG wireframes, flow map, and published HTML viewer from the Artifacts page.
Outcome: A reviewable multi-screen wireframe lands in the task as an artifact you can annotate and share instead of a text answer.
Use Cases
- Orchestrate AI agents for development tasks like coding, testing, and deployment
- Run autonomous content creation, social media management, and marketing campaigns
- Set up QA and research teams that operate independently
- Manage outreach, sales, and customer support with a multi-agent crew
- Use the org chart to assign budgets and goals across departments
- Prototype an agent team quickly via `npx paperclipai onboard` or the self-serve sandbox
- Financial operations: an agent can incorporate, apply, and stand up a finance stack via Ramp
- Turn a one-line prompt into a reviewable multi-screen wireframe flow you can share
Models Under the Hood
as of 2026-10-17
Limitations
- Paperclip is open-source and self-hosted: your agents, your data, your infrastructure, so you run and maintain your own deployment (laptop, Docker, or server).
- Onboarding and administration lean on the CLI and API for local setup, identity, instance configuration, and admin tasks, which implies a learning curve for non-terminal users.
- The release cadence is aggressive and several 2026 releases are flagged breaking (including v2026.1012.0 on October 12, 2026), so staying current means reading upgrade guides.
- A hosted version currently appears to exist only as a waitlist.
as of 2026-10-04
Verification history
We have re-verified Paperclip 10 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 10 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Paperclip tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Open Source
$0/mo
Ideal for
Technical builders and small teams who are willing to self-host via `npx paperclipai onboard` and pay only their own infrastructure and model-provider usage.
What this tier adds
Starting tier: the full MIT-licensed platform, including the org chart, budgets, approvals, Connectors, and Artifacts page.
Where the pricing makes sense
The company stage and team size where Paperclip's pricing actually pencils out — and where peers do it cheaper.
Paperclip is open source under the MIT License, with a self-hosted local install available now and a hosted version on a waitlist. That puts the entry cost at your own infrastructure plus model-provider usage rather than a per-seat subscription — cheaper than managed orchestration services if you have the engineering time, more expensive in hours if you don't.
Setup time & first value
How long it actually takes to get something useful out of Paperclip — broken out by persona, not the marketing-page minute.
Solo technical builder: roughly an afternoon from `npx paperclipai onboard` to a working org chart with your first agent and budget cap, plus time to connect GitHub or your runtime adapter. Small team: add a session to wire Connectors, per-user secrets, and the approval queue so more than one person is on it. Non-technical operators: treat it as days, not hours, because you'll be working in a
Switching to or from Paperclip
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From a pile of standalone agent scripts: install locally with `npx paperclipai onboard`, then recreate each script as an agent in the org chart with its own budget and task queue.
- →From Claude Code or Codex session runners (Agent View, Agent Teams, amux): connect the same runtimes via adapters and gain budgets, approvals, and an audit trail around them.
- →From chat-based prompting: move recurring work into goal-backed tasks and routines so agents pick up the next item instead of waiting on a prompt.
- ↗To a managed orchestration service: export your company data and agent configuration, then rebuild the org chart in the hosted product once you're ready to leave self-hosting behind.
- ↗To raw CLI agents: keep your adapters and runtimes (Claude, Codex) and drop the governance layer, accepting that budgets and approvals go away with it.
- ↗To a single-session runner: if you only ever need one agent at a time, a lightweight session manager is less machinery than an org chart.
Integrations
Resources & Guides
Tutorials & Learning
YouTube returned 6 videos for “Paperclip”, and we withheld 6: 6 could not be judged, because “Paperclip” is a single word that other videos use for other things. We are showing none, because we could not prove any of them are about Paperclip.
Official links
Featured Head-to-Head Comparisons
Paperclip vs Locus Robotics
Locus Robotics and Paperclip serve completely different domains: physical warehouse automation vs. digital AI agent orchestration. Choose Locus if you run a high-volume fulfillment center and need proven AMRs to boost productivity 2-3x, but be ready for a custom-priced RaaS commitment. Choose Paperclip if you're a developer building autonomous AI teams for code, content, and outreach, and want a free, open-source management UI. There's no overlap—your use case decides.
Paperclip vs Presto Voice
Presto Voice is the clear choice for QSR chains seeking to automate drive-thru ordering and boost revenue through proven upselling, especially with its recent Dairy Queen partnership. Paperclip is ideal for developers and builders who want to orchestrate autonomous AI agent teams for business tasks, but it requires self-hosting and technical setup. Choose Presto if you run a drive-thru; choose Paperclip if you want to manage a company of AI agents.
Paperclip vs Truleo
Truleo and Paperclip serve entirely different markets. Truleo is a specialized law enforcement intelligence platform that connects siloed data (RMS, CAD, jail calls, BWC) for automated lead generation and report writing. Paperclip is an open-source, freemium platform for developers to build and manage autonomous AI agent teams. Buyers should choose based on their domain: Truleo for police agencies, Paperclip for tech builders.
Popular in Agent Frameworks & Orchestration
Temporal AI
Durable execution platform that keeps AI agents and long-running workflows alive through crashes, retries, and abandoned sessions.
Frequently Asked Questions
Used Paperclip? Help shape our editorial sentiment research.