⚠The team behind this tool is not publicly identified. Treat with extra caution — anonymous authorship is a significant trust risk.
Paperclip
Agent Orchestration
D
Overall score: 2.3Reviewed April 14, 2026
Self-hostable and MIT-licensed at the core, but a fully pseudonymous operator with no legal accountability, a ToS granting irrevocable ML training rights over submitted data, and a six-week-old codebase make this unsuitable for production or any sensitive use.
Score Summary
Claim Accuracy3/5
Data & Privacy2/5
Security Posture3/5
Transparency1/5
Key Findings
›The sole founder operates pseudonymously as '@dotta' with no verifiable legal name on the official website or documentation; Paperclip Labs, Inc. is the named legal entity in the ToS (paperclip.ing/tos) but lists no executives, founders, or investors publicly (source: paperclip.ing, paperclip.ing/tos).
›ToS Section 6 (paperclip.ing/tos) grants Paperclip Labs a worldwide, royalty-free, transferable, sublicensable licence over all submitted Content explicitly for ML training and commercialisation; this licence survives account termination with no described opt-out for already-submitted data.
›ToS Section 7 (paperclip.ing/tos) further grants ML training and commercialisation rights over opt-in detailed telemetry including agent run logs and configuration data, with no post-consent revocation mechanism described.
›Anonymous usage telemetry is enabled by default in self-hosted installs and must be explicitly disabled via PAPERCLIP_TELEMETRY_DISABLED=1, DO_NOT_TRACK=1, or config file (source: github.com/paperclipai/paperclip README).
›Project launched March 4, 2026 — approximately six weeks old at review time — with 5 releases, 686 open GitHub issues, and no documented enterprise adoption or third-party security audit (source: github.com/paperclipai/paperclip).