dsh-codebase-memory-mcp
Run the following command in DeepSeek Harness:
dsh plugin install princeofdream/dsh-codebase-memory-mcp
Paste the following prompt into your AI chat to install this plugin:
Run dsh plugin install princeofdream/dsh-codebase-memory-mcp inside DeepSeek Harness to install the plugin; the source is available at https://github.com/princeofdream/dsh-codebase-memory-mcp . Reload the host after installation to activate the tools.
About this plugin
When working with a local codebase inside DeepSeek Harness, the model is often constrained by context length and cannot trace call chains across files, query structural relationships, or assess the blast radius of a change. dsh-codebase-memory-mcp bridges the local codebase-memory-mcp knowledge-graph server into DSH through the built-in @deepseek-ai/dsh-mcp-client layer, giving the model direct access to 14 graph-aware tools including index_repository, search_graph, trace_path, query_graph, and detect_changes, effectively turning the repository into a queryable, traceable knowledge graph.
On top of the bridge itself, the plugin adds a practical operational layer: the settings namespace exposes five hot-reloadable knobs (exePath, cwd, extraArgs, env, enabled) so configuration changes take effect without restarting the host. Two host-facing tools, codebase_memory_status and codebase_memory_restart, let either the user or the model inspect bridge state (idle / connecting / ready / error / disabled) at any time and re-launch the server after fixing a misconfiguration, preventing hard crashes when the binary cannot be resolved.
This plugin is built for DSH power users who work daily against a local repository and want the model to reason over project structure, trace cross-file dependencies, and detect uncommitted changes grounded in a graph. No custom MCP transport code is needed; once installed, the knowledge-graph capability is available in the workflow out of the box.
Use Cases
- Trace function call chains and module dependencies across files
- Query code structure via knowledge graph to answer architecture questions
- Detect uncommitted changes and assess their blast radius
Best For
- Developers who work daily against local repos in DSH
- Teams that need the model to grasp overall project structure
- Engineers who want to embed code-graph reasoning into their AI workflow
Related Plugins
Traceable, searchable cross-session memory for AI agents that turns conversation knowledge into a typed knowledge graph and recalls relevant subgraphs instead of replaying full history, natively integrated with DeepSeek Harness.
Gives DSH AI cross-session long-term memory, to-do and skill management, plus multi-session orchestration, external AI delegation, and an infinite canvas that grows with you.
dsh-mnemon is a three-tier, pluggable, Agent-driven memory system for DeepSeek Harness, combining Runtime memory, Project Documents, and replaceable Memory Spaces with nine long-term providers.
An opinionated, zero-infrastructure file-based memory protocol: human-readable markdown files plus a bounded always-loaded index, curation discipline, and no database or embeddings.