?? DESCRIPTION

EvoMap accused Hermes Agent of copying Evolver’s architecture, citing over a dozen technical indicators, including memory systems and reflection mechanisms. Nous Research dismissed the claims, with founder Teknium calling them “completely brainless.” The dispute underscores strong interest in AI agent development, as Hermes Agent has 88,000 GitHub stars compared to Evolver’s 2,000. EvoMap insists on attribution, while Teknium refuses to engage on technical grounds.

ME News reports that on April 15 (UTC+8), according to monitoring by Beating, EvoMap, the development team behind the AI Agent self-evolution engine Evolver, published a technical comparison article on April 14 accusing Nous Research’s open-source project Hermes Agent of systematically replicating Evolver’s architecture at the structural level, without any attribution in any publicly available materials. Hermes Agent currently has 88,000 GitHub stars, while Evolver has approximately 2,000. Teknium, founder of Nous Research, responded on X stating, “I’ve never heard of this person or this project,” and called the comparison article “completely brainless.” The official Nous Research account replied even more briefly: “Our repository was created in July 2025. We are pioneers of the underlying technologies for modern Agent frameworks like YaRN. Delete your account.”EvoMap’s article lists over ten architectural correspondences module by module: 1. A three-layer memory system (persistent facts + procedural knowledge + historical search), with each layer’s function and role precisely aligned. 2. A closed-loop mechanism for automatically extracting reusable experience assets after task completion (Evolver calls them Gene/Capsule; Hermes calls them SKILL.md). 3. A periodic reflection mechanism (Evolver triggers every 5 evolution cycles; Hermes triggers every 15 tool calls). 4. Runtime skill discovery and on-demand loading. 5. A 10-step orchestrated evolution cycle. 6. Engineering patterns including multidimensional weighted scoring, constraint gating, and atomic writes.On the timeline: Evolver’s core protocol GEP was fully released between February 1–16, with 136 total releases. Hermes’ skill system was only released on March 12 (v0.2.0), and its self-evolution repository was created on March 9—24 to 39 days after GEP’s public release.However, the article also acknowledges facts favorable to Hermes: The Hermes Agent repository was created in July 2025—six months before Evolver—but remained private until late February; Hermes’ self-evolution module uses the academic framework GEPA from Stanford/Berkeley (ICLR 2026 Oral), unrelated to GEP; Anthropic’s Agent Skills standard, released in December 2025, predates both projects. The two projects use different languages (JavaScript vs. Python), and no code-level copying has been found.Individually, any one of these dimensions could easily result from independent convergence: three-layer memory, experience extraction, and periodic reflection are widely discussed topics in the AI Agent field. EvoMap’s argument rests on probabilistic reasoning—that the simultaneous alignment across more than ten dimensions exceeds coincidence. This argument is compelling but not conclusive. Teknium chose to deny knowledge rather than address the technical details, offering no further explanation. This dispute is unlikely to be resolved in the short term, but it highlights an ongoing issue: when asserting architectural precedence against a vastly more influential open-source project, the smaller party faces immense challenges in both evidence presentation and public communication.(Source: BlockBeats)Source:Show originalDisclaimer: The information on this page may have been obtained from third parties and does not necessarily reflect the views or opinions of KuCoin. This content is provided for general informational purposes only, without any representation or warranty of any kind, nor shall it be construed as financial or investment advice. KuCoin shall not be liable for any errors or omissions, or for any outcomes resulting from the use of this information.Investments in digital assets can be risky. Please carefully evaluate the risks of a product and your risk tolerance based on your own financial circumstances. For more information, please refer to our Terms of Use and Risk Disclosure.