Dimension scores are derived from public data and fields; weighted into the composite. Reference only.
Paperclip is an open-source AI Agent orchestration app from Paperclip Labs, Inc. Its goal is not to replace Claude, Codex, or Cursor, but to organize those agents into a “company-like team.” Users first define business objectives, then configure roles such as CEO, CTO, engineers, and marketing, and run work through an org structure, tickets, budgets, and approvals.
Its biggest differentiator is Bring Your Own Agent: any agent, script, shell command, HTTP webhook, or runtime such as Claude Code/OpenClaw that can receive a heartbeat can be connected. The system provides an org chart, goal alignment, automatic delegation, scheduled wake-ups, ticket tracking, full tool-call traces, and immutable audit logs. Cost controls are also fairly complete: monthly budgets can be set per Agent, with alerts at 80% and automatic pausing at 100%, making it useful for preventing multiple Agents from burning through tokens uncontrollably.
The main materials do not disclose pricing for a commercial hosted version. The open-source project is MIT-licensed, can be self-hosted, and does not require a Paperclip account. Locally, it can run as a single Node.js process maintaining an embedded Postgres instance, or it can point to your own Postgres database or a remote deployment. Actual costs mainly come from the LLMs, Agents, cloud resources, and external tools that users connect.
The advantages are that it is model- and Agent-neutral, extensible, self-hostable, and transparent in auditing, while elevating Agent collaboration from a “prompt-tool” approach to a “team governance” model. The drawbacks are also clear: it is closer to engineering infrastructure, so users need to handle databases, authentication, adapters, and Agent permissions themselves. What Agents can do with external systems also needs to be hardened by the user. On privacy, self-hosting helps with data control, but its terms state that hosted services collect anonymous telemetry; after opting in to detailed telemetry, this may include runtime logs and configuration, and may be used for product and model improvement.
Paperclip is suitable for technical founders, AI automation teams, and scenarios that require multi-Agent collaboration, such as software development, content marketing, research, and outreach. It is less suitable for non-technical users who just want something ready to use out of the box. Access from China, payment options, and Chinese-language support are not specified in the main materials. Availability of the website and the GitHub/npm ecosystem may depend on the network environment. Alternative or complementary tools include OpenClaw, Claude Code, Cursor, Codex, as well as task systems such as Asana/Trello.
⚠ This review is compiled from public sources and does not constitute a purchase recommendation. Verify all facts on the vendor's official site. Verify on paperclip.ing official site.
paperclip.ing is an Unknown AI Apps provider. TG4G tracks its product information, an overall rating of 8.0/10, and a China-accessibility score of China direct-connect friendly. Click "Visit Official Site" to reach paperclip.ing directly.