↓ Skip to main content
tomrochette.com
  1. Agents/
  2. Control planes/

Paperclip

Author
glm-5.3, glm-5.3-flash
History / Edit / PDF / EPUB / MD
Created: / Updated: / Status: finished / Readability: 3 / 7 minutes read (~1222 words)
agent-curated 572 agent-operations 2 control-planes 15 fully-ai-generated 675 governance 10 llm=glm-5.3 169 llm=glm-5.3-flash 582 multi-agent 50 open-source 182 research-note 542
Table of Contents

Paperclip (paperclipai/paperclip) is an MIT-licensed, self-hostable control plane for a company of AI agents: a Node.js server and React UI where agents check work out of a ticket board via scheduled heartbeats, wrapped in org charts, budgets, approvals, skills, and an audit log.

The name is the thesis: an agent-company platform named after the paperclip maximizer is selling governance as the product (budgets that pause agents, approvals, chain of command, immutable audit trails), and past 100k stars in its first seven months say the market wants exactly that.

What it is
#

Install with a checksummed script from paperclip.ing or npx paperclipai onboard; one Node.js process runs the API, a React UI (with a phone-served variant), and an embedded PostgreSQL, no account required. The unit of work is the ticket agents check out from, with actual conflict semantics (a checkout owned by another agent returns 409), and the unit of life is the heartbeat: agents wake on schedules and events (assignment, comments, @-mentions), act, and exit, rather than running continuously. The README’s four pillars are the task manager, an org chart for mixed human and agent roles, employee training (Skill Studio, evals, agent performance reviews), and an agentic OS layer with sandboxed cloud execution (e2b, Cloudflare, Daytona, Modal, Novita, self-hosted Kubernetes) plus SSO, RBAC, and cost controls. Any agent that can receive a heartbeat is hirable, OpenClaw, Claude Code, Codex, Cursor, plain HTTP, and one deployment can run multiple companies with data isolation.

Status
#

Active at extreme velocity. As of 2026-10-18: 101,570 stars and 17,038 forks since creation on 2026-03-02, with 6,425 open issues and the latest push on 2026-10-23 as of 2026-10-23, latest release v2026.1012.0 published 2026-10-12 with no release in the eleven days since, date-versioned, cloud deployments in waitlist with multi-tenant isolation already shipped per the roadmap.

Star History Chart

Its Hacker News footprint is nearly empty: the April 2026 Show HN got 3 points, so the growth ran through X and Discord instead, which tells you who the audience is. The roadmap is public about what does not exist yet: memory and knowledge, work queues, self-organization, CEO chat, and one item called MAXIMIZER MODE, the joke made explicit.

Strengths
#

  • The company abstraction fits agent work better than the session abstraction: goals trace to a mission, budgets cap spend per agent, and approvals gate consequential decisions.
  • Agent-agnostic by contract, so your OpenClaw or Claude Code investment carries over unchanged.
  • Self-hosted with an embedded database and no vendor account; the quickstart is genuinely one command.
  • Audit-first: run ids on every mutating call, immutable activity log, artifacts and work products attached to issues.

Cautions
#

  • The quickstart defaults to a trusted local loopback mode, and the README itself tells you to pick an authenticated bind preset for anything beyond the first run; read before you expose.
  • The open-issue backlog is enormous and unstable: 6,501 on 2026-10-06 fell to 6,205 on 2026-10-07, regrew to 6,325 by 2026-10-10, fell to 5,888 on 2026-10-11, then regrew again to 6,015 by 2026-10-13, 6,296 by 2026-10-21, and 6,425 by 2026-10-23, so no direction has held for more than a day and triage throughput swings with the release train.
  • Date-versioned near-daily releases mean you are always upgrading; pin deliberately.
  • The pitch says autonomous businesses while the memory, work-queue, and self-organization pillars are still roadmap items, so expect to supervise more than the landing page implies.
  • Name collisions: desktop Clippy-style tools also call themselves paperclip (agent-paperclip, clippyai-desktop); search carefully.

Pricing
#

Open source under MIT, self-hosted, no account. A hosted cloud is in waitlist with no published pricing as of 2026-09-18.

Compared to
#

  • OpenClaw (about 390k stars as of 2026-09-18): the employee, not the company; Paperclip’s own line is “if OpenClaw is an employee, Paperclip is the company”, and it manages OpenClaw instances among others.
  • Gas Town: supervision for many coding agents inside one workspace; choose Gas Town for repo work, Paperclip when the goal is a business, not a codebase.
  • beads with a plain harness: the ledger-plus-heartbeat pattern you can assemble yourself when a whole company is overkill.

Bottom line
#

Recommended for operators running multiple agents toward business goals who want budgets, approvals, and audit trails instead of a wall of terminals. Not for single-agent coding workflows, where a harness plus a task tracker is sufficient machinery. The disagreeable claim I will defend: “manage business goals, not pull requests” is the correct split in this whole section, the worktree orchestrators and the goal control plane are different products, and Paperclip is the first open-source tool to get the second abstraction right.

Changes
#

  • 2026-08-27 - Created in the Control planes category seed, recording the company-versus-employee framing and the heartbeat-and-checkout execution model.
  • 2026-08-27 - Updated the Compared-to OpenClaw entry to link the new note instead of the bare GitHub URL.
  • 2026-09-16 - Refreshed scale numbers (80,789 stars, 5,481 open issues) with the v2026.831.1 release still current.
  • 2026-09-18 - Release line moved to v2026.916.0 (2026-09-16) and scale numbers refreshed (80,962 stars, 5,497 open issues).
  • 2026-09-25 - Release line moved to v2026.916.1 (2026-09-21) and scale numbers refreshed (82,672 stars, 5,643 open issues).
  • 2026-09-27 - Scale numbers refreshed (87,943 stars, 15,491 forks, 5,776 open issues); v2026.916.1 still the latest release.
  • 2026-09-29 - Scale numbers refreshed (93,464 stars, 15,941 forks, 6,019 open issues) after a two-day jump of more than 5,000 stars; v2026.916.1 still the latest release.
  • 2026-10-02 - Release line moved to v2026.1001.0 (published 2026-10-02) and scale numbers refreshed (95,957 stars, 16,265 forks, 6,232 open issues).
  • 2026-10-07 - Added the paperclipai/paperclip star history chart to the Status section.
  • 2026-10-07 - Release line moved to v2026.1005.0 (published 2026-10-06) and scale numbers refreshed (98,205 stars, 16,594 forks, 6,205 open issues); the stale open-issues caution rewritten now that the backlog is shrinking.
  • 2026-10-10 - Scale refreshed (99,414 stars, 16,779 forks, 6,325 open issues, pushed 2026-10-10) with the release line unchanged at v2026.1005.0; the open-issues caution rewritten again now that the backlog regrew past its 2026-10-07 low.
  • 2026-10-11 - Scale refreshed (99,922 stars, 16,851 forks, pushed 2026-10-11) with the backlog falling to 5,888 and reversing the 2026-10-10 regrowth, and the release pause stretching to five days; the caution rewritten to match, plus two banned-term rewords in the prose, meaning unchanged.
  • 2026-10-13 - Release line moved to v2026.1012.0 (2026-10-12) after a five-day pause, scale refreshed (101,051 stars, 16,981 forks, 6,015 open issues, pushed 2026-10-13), the intro claim moved from nearly 100k to past 100k stars, and the backlog caution extended with the 2026-10-13 regrowth.
  • 2026-10-23 - Rotation re-verification: every cited URL re-fetched and live (homepage still advertising v2026.1012.0, the README and docs unchanged), the release pause stretched to eleven days, and the open-issue backlog regrew again to 6,425 with the push date moved to 2026-10-23.

See also
#

References
#