跪拜 Guibai
← All articles
Artificial Intelligence · Agent

DeepSeek's Cordis Protocol Turns Agent Development into a Lego Set

By 一拳不是超人 ·
Read original on juejin.cn ↗ Google Translate ↗ Alt translation

Agent frameworks today lock developers into walled gardens where core behavior can't be changed without forking, and plugins written for one framework are useless in another. A protocol-first standard under MIT license, backed by a major model provider and already proven across 4,000 community plugins, could break those walls and make agent components genuinely reusable across projects.

Summary

DeepSeek's Harness and Cordis release replaces monolithic agent frameworks with a protocol layer where every component, from the agent loop to tool integrations, is an ordinary plugin that can be swapped without forking. The protocol governs service registration, dependency declaration, side-effect disposal, event communication, and component lifecycle—no part of the system holds special status.

Cordis has already been validated at scale through Koishi, a chatbot framework with over 4,000 community plugins that interoperate purely through the protocol. The same protocol can assemble a customer-service agent with ticketing and knowledge-base plugins, a code-review agent with Git and CI plugins, or a security-operations agent with log analysis and auto-response—all sharing the same plugin marketplace.

The paper also outlines a self-evolving agent harness: because Cordis supports reversible effects and hot module replacement, an agent can unload and replace its own components at runtime without losing state or interrupting service. The team acknowledges this remains a research direction rather than a shipped product capability, with validation limited to the TypeScript-only Koishi ecosystem.

Takeaways
DSH (DeepSeek Harness) is an agent orchestration layer released under MIT license alongside Cordis, a meta-framework that defines protocols rather than a fixed architecture.
In Cordis, the agent loop, tool scheduler, and state management are ordinary plugins—swap them by registering a new plugin under the same service name, no fork required.
Koishi, a chatbot framework built on Cordis, has accumulated over 4,000 community plugins over four years, all interoperating through the shared protocol.
The same Cordis baseplate can produce a customer-service agent, a code-review agent, a security-operations agent, or even a non-agent microservice gateway.
Cordis's reversible effects allow runtime component replacement: unload triggers automatic LIFO rollback of all registered side effects, then load resolves dependencies and registers new ones.
The paper proposes self-evolving agents that can automatically load or unload plugins based on operational experience—for example, detecting frequent Japanese queries and loading a translation plugin.
The authors acknowledge that self-evolving agent runtimes remain a research direction, with validation currently limited to the TypeScript-only Koishi ecosystem and no controlled comparisons against alternative architectures.
Conclusions

DeepSeek is betting that agent infrastructure, not model capability, is the bottleneck—a position that puts it in direct competition with framework vendors like LangChain rather than model providers.

By releasing Cordis under MIT, DeepSeek is making a standards play: if the protocol becomes a de facto standard, plugin portability across projects and teams becomes the norm, and existing walled-garden frameworks lose their lock-in advantage.

The self-evolving agent concept is less about AI magic and more about deterministic engineering—reversible side effects, reactive dependency graphs, and hot module replacement—which means it's auditable and debuggable in ways that purely model-driven approaches are not.

The 4,000-plugin Koishi ecosystem is the strongest signal that a protocol-first approach works at scale, but it's a single-language, single-domain proof point; cross-language and cross-domain adoption remain unproven.

Concepts & terms
Cordis
A meta-framework that defines protocols for service registration, dependency declaration, side-effect disposal, event communication, and component lifecycle. Unlike traditional frameworks, it has no privileged core code—every component, including the agent loop, is an ordinary replaceable plugin.
Reversible Effects
A mechanism in Cordis where every side effect a plugin registers (event listeners, services, sub-plugins) is tracked and can be automatically rolled back in LIFO order when the plugin is unloaded, enabling safe runtime component replacement without state loss.
Self-Evolving Agent Harness
A proposed system where an agent can automatically generate, load, and unload its own plugins at runtime based on operational experience, without interrupting service. Enabled by Cordis's reversible effects and hot module replacement.
Koishi
A chatbot framework built on Cordis that has accumulated over 4,000 community plugins over four years, serving as the first large-scale validation of the protocol-first approach.
Source: juejin.cn ↗ Google Translate ↗ Backup ↗