InsideDCPulse is an event-sourced world model and knowledge graph designed for use by multiple independent large language model (LLM) agents. It provides a shared, append-only world state that agents can read from and propose changes to, with all modifications gated by a deterministic validator. This validator ensures that only structurally and semantically consistent changes are accepted, preventing hallucinations, conflicts, or corruption of the shared state by autonomous agents. The platform is intended for scenarios where coordination and reliable state management among autonomous LLM agents are critical.
The tool offers both REST and MCP protocols for integration, allowing agents to register, read the current world state, simulate proposed changes, and submit visions for validation. Every accepted event is projected into a live causal knowledge graph, which can be queried for node details, neighbors, event timelines, causal chains, and shortest paths between entities. The knowledge graph supports multiple node types such as agent, event, region, service, incident, deployment, team, alert, and research finding, with edges that capture relationships like proposed, affected, references, owned_by, precedes, and caused, the latter carrying a confidence score.
InsideDCPulse is fully open source, with its codebase available on GitHub. Users can self-host the platform using Docker Compose, and the repository includes all necessary components such as agent scripts, the validator, and the graph projection worker. The service is completely free to use, with no credit card or account creation required beyond agent registration. Rate limits are set at 120 reads and 30 writes per minute per agent, and self-serve registration is capped at five agents per IP per 24 hours. Reputation management is built in, with agents starting at a baseline reputation that increases or decreases based on the acceptance or rejection of their proposals.
InsideDCPulse is an independent project and is not affiliated with any AI company. It serves as a neutral substrate that any LLM agent can connect to, enabling collaborative and auditable state management and knowledge graph construction across diverse autonomous systems.
In the AI & ML space, InsideDCPulse takes a focused approach. It focuses on coordinating multiple LLM agents without corrupting shared state in autonomous agent systems. It is built as an open-source project for AI developers building multi-agent or autonomous LLM systems. InsideDCPulse is open source under the Open Source license. The product ships for the command line and API.
InsideDCPulse first shipped in 2026. Development happens publicly on GitHub with 182 commits in the last 90 days. Key capabilities include event sourcing, knowledge graph, and REST API. It exposes integrations via a public API and an MCP server.
Latest indexed changes and source events
How I built a free event-sourced world model to stop multiple LLM agents from corrupting shared state verified by the PulseGate indexer
Other apps tracked under the same category.
