Components you configure, not a hosted offering
“Services” here means the software components referenced by the Dopemux repository. This site does not advertise consulting, managed hosting, a cloud subscription, or an automatically provisioned account.
A usable local setup depends on the components you select, their storage, credentials, transports, and your client configuration. The directory explains what source and configuration establish, with limitations on each detail page.
Explore the components
Search by name, role, or capability. Status labels describe source evidence, not current runtime health. Read the status legend before treating a configured entry as available.
15 entries
ConPort
Structured project context, decisions, and progress: a place to retain the reasoning around the work, separately from the work’s historical event stream.
Explore ConPortDope-memory
An inspectable chronicle of work: activity records, recaps, session replay surfaces, and links between earlier events and later resolutions.
Explore Dope-memoryDope-context
Code and documentation retrieval for a workspace, with explicit indexing operations kept separate from normal search exposure.
Explore Dope-contextADHD Engine
A configured operator-state service for recommendations and accommodation-related signals. Its role is experimental support, with no clinical claim.
Explore ADHD EngineRepo Truth Extractor
Plan repository evidence extraction through a canonical CLI, with explicit phase, provider, routing, and run controls.
Explore Repo Truth ExtractorDopecon-bridge
Routing, proxying, and event transport between separate systems. A connection layer with a deliberately bounded role.
Explore Dopecon-bridgeTask Orchestrator: two distinct surfaces
Understand the workflow MCP integration and the separate Python coordination API before selecting a connection or interpreting its state.
Explore Task Orchestrator: two distinct surfacesLeantime & its PM bridge
A project-management application and adapter for projects, tickets, and milestones. PM metadata remains a distinct part of the workflow.
Explore Leantime & its PM bridgeLiteLLM proxy
A configured local model-access endpoint with aliases, routing, retries, and fallbacks. Inspect the actual route before relying on its label.
Explore LiteLLM proxyQdrant
Vector-storage infrastructure supporting semantic retrieval indexes. It is a backing dependency, not automatically an assistant-facing MCP server.
Explore QdrantSerena
A code-intelligence integration intended for symbol and code-relationship navigation. Its current transport declarations need qualification.
Explore SerenaPAL reasoning tools
A configured tool surface for structured analysis, planning, review, and related reasoning workflows. Distinguish its stdio tools from the HTTP health shim.
Explore PAL reasoning toolsResearch tooling
Configured research and documentation integrations for evidence gathering, with external credentials and client exposure kept explicit.
Explore Research toolingRead-only evidence facade
Six local target-evidence tools for inspecting capabilities, repository snapshots, proof bundles, and runtime-receipt evidence.
Explore Read-only evidence facadePostgreSQL & Redis
Structured persistence, caches, and event streams beneath the tools. Different storage jobs remain separate, even when they share a deployment.
Explore PostgreSQL & RedisNo matching entries. Clear the search or choose another status.
Choose the layer that matches the question
For a project decision, look to ConPort. For a history of work, look to dope-memory. For a code passage, look to dope-context. For PM metadata, distinguish Leantime from the workflow-transition tooling and external execution runtime.
Backing stores are included so the system is easier to understand, but databases and caches are not automatically MCP servers. The server directory organizes the same components by connection role.
Source evidence
Inspected source at the pinned reference revision. Source presence does not establish runtime health.