Усильте агента
Собственная команда издателя. Не эндпоинт PlugMCP.
Host a copy with our keys and logs.
npx -y @lovedolove/[email protected]У Claude Code нет ссылки для установки локальной команды. Скопируйте команду ниже и выполните её в терминале.
claude mcp add lovedolove-dsh-codebase-memory-mcp -- npx -y @lovedolove/dsh-codebase-memory-mcp@0.1.10Выполните в терминале.
{ "mcpServers": { "lovedolove-dsh-codebase-memory-mcp": { "command": "npx", "args": [ "-y", "@lovedolove/[email protected]" ] } }}claude_desktop_config.json
Удалённый URL или команда на вашем компьютере? Чем различаются два способа установки
Запустить на PlugMCP
A new server you create. It is not this listing's official endpoint. Use the publisher install when you want their URL and their credentials.
ЗапускОт издателя
Standalone Codebase Memory MCP bridge plugin for DeepSeek Harness (DSH). Registers six knowledge-graph code tools, auto-starts the
codebase-memory-mcpdaemon with its graph-visualization Web UI, and ships a bundled agent skill plus a/cbmstatus command.
Codebase Memory MCP bridge plugin for DeepSeek Harness: knowledge-graph code search, snippets, architecture, call-path tracing, and auto-started graph visualization UI.
cbm_projects, cbm_search, cbm_snippet, cbm_arch, cbm_trace,
cbm_search_code). The bundled skill's guidance notes that graph queries return
precise structural results in ~500 tokens vs ~80K for grep.codebase-memory-mcp --ui=true as a child process over stdio JSON-RPC 2.0
(MCP handshake 2025-03-26) when the plugin loads, and prints
codebase-memory UI is live: http://localhost:9749/. The shared client
restarts the daemon automatically if it exits and disposes it when the plugin
shuts down.grep/read.skills/codebase-memory/SKILL.md
is registered as a native bundled skill provider (codebase-memory, rank 600,
model- and user-invocable), so the model-facing skill tool can load it./cbm slash command — reports the resolved executable path, the Web UI
link, and the projects currently indexed in the daemon.tools is a hard dependency (inject: ['tools']);
systemPrompt, skills, and commands are attached via deferred ctx.inject(),
so a missing host service simply skips that feature.CBM_EXE override, user and system
directories, current working directory, PATH, and a WSL interop fallback that
locates codebase-memory-mcp.exe reachable from Linux/WSL.Optional peer dependencies (declared in peerDependencies, resolved from the DSH
runtime when present): @deepseek-ai/cordis >=4.0.1-rc.1 <5.0.0-0, and
@deepseek-ai/dsh-commands, @deepseek-ai/dsh-skill,
@deepseek-ai/dsh-system-prompt, @deepseek-ai/dsh-tools >=0.1.2-rc.1 <0.3.0-0.
Install the executable before booting DSH. The plugin resolves the binary when it loads; if none is found, the daemon is not started and the
cbm_*tools are not registered. Installing the binary later requires reloading the profile.
codebase-memory-mcp executableThe plugin is only a bridge — code indexing and querying are done by the
codebase-memory-mcp runtime (upstream project,
documentation):
# Installs the CLI and downloads/caches the native runtime for your platform
npm install -g codebase-memory-mcp
# Verify it runs
codebase-memory-mcp --version
Alternatively, download your platform's release binary and place it in one of the locations the plugin searches:
~/.local/bin/codebase-memory-mcp (Linux/macOS user bin)/usr/local/bin/ or /usr/bin/ (system directories)PATH%USERPROFILE%\AppData\Local\Programs\codebase-memory-mcp\codebase-memory-mcp.exe (Windows)Or point the plugin at it explicitly with CBM_EXE=/path/to/codebase-memory-mcp.
dsh plugin --profile web add @lovedolove/dsh-codebase-memory-mcp
(replace web with your profile name — web, headless, sdk, acp, …)
This forwards the package install into the profile directory. Because the
package declares dsh.bundle.patch, DSH automatically activates the bundle
by appending @lovedolove/dsh-codebase-memory-mcp to the profile's
dsh.profile.bundles list.
Other ways to install:
Web sidebar → Plugins page: search for the package and install it there
(the plugin_manager tool exposes the same operations in Creator mode).
Manual wiring: add the package to the profile's dependencies in
$DSH_HOME/profiles/<name>/package.json and list it under
dsh.profile.bundles, then run a package install in that directory. The
shipped bundle patch (cordis.patch.yml, declared as
dsh.bundle.patch) then mounts the plugin:
- insert:
- id: codebase-memory
name: '@lovedolove/dsh-codebase-memory-mcp'
On plugin load you will see:
[codebase-memory] 🔍 codebase-memory UI is live: http://localhost:9749/
and, in the log:
[codebase-memory] registered tools: cbm_projects, cbm_search, cbm_snippet, cbm_arch, cbm_trace, cbm_search_code
[codebase-memory] registered system prompt guidance
[codebase-memory] registered /cbm command
[codebase-memory] registered skill: codebase-memory
The plugin injects the knowledge-graph policy into the system prompt and bundles
the codebase-memory skill, so the agent picks the right tool on its own.
Typical workflow:
cbm_projects() — confirm the project is indexed (by default the project is
inferred from the workspace).cbm_search(name_pattern=".*auth.*") — find symbols by name/label/path.cbm_trace(symbol="login", direction="both", max_depth=3) — trace callers
and callees.cbm_snippet(qualified_name="app.auth.login") — read the exact definition
without loading whole files.cbm_arch(directory="src") — get a directory-level component overview.Quick decision matrix (from the bundled skill):
/cbm🔍 **Codebase Memory MCP**
• Executable: `/home/you/.local/bin/codebase-memory-mcp`
• Web UI: http://localhost:9749/
**Indexed Projects:**
```
home-you-my-repo
```
If the binary is missing, /cbm reports:
⚠️ codebase-memory-mcp executable not found. Install codebase-memory-mcp or set CBM_EXE.
The daemon is launched with --ui=true, so an HTTP graph-visualization UI is
served at http://localhost:9749/ (the runtime's default port; the upstream
binary accepts --port=N and persists it — the plugin's prompts and /cbm
assume the default 9749).
All six tools accept project (explicit project slug) and/or repo
(repository root path, slugified automatically) and are marked concurrency-safe.
Results are returned as text.
Each tool forwards to the daemon's MCP tools/call method
(list_projects, search_graph, get_code_snippet, get_architecture,
trace_path, search_code) with a 120-second timeout; tool-level errors are
surfaced as failed tool results.
When a tool is called, the project is resolved as:
project argument,repo argument,CBM_DEFAULT_PROJECT,Slug mapping: /home/user/my-repo → home-user-my-repo; Windows paths keep a
drive prefix, e.g. C:\Users\me\app → C-Users-me-app.
The daemon starts eagerly when the plugin applies (unless the process runs
under node --test), and every tool call ensures it is running, so the child
process is restarted transparently if it dies. On plugin disposal the daemon is
killed together with any pending requests.
# Unit + integration test suite (node --test)
npm test
# Verify npm package packing
npm run pack:check
The suite covers path slugification, findExe/CBM_EXE lookup, the stdio
JSON-RPC client (spawn, handshake, tool calls, errors), tool registration, a full
apply() against a mocked Cordis context, and — when the real binary is
installed on the host — a live list_projects round trip (this test is skipped
otherwise). CI runs the suite on Ubuntu with Node 20, 22, and 24.
Repository layout:
src/index.mjs # Cordis plugin: apply(), system prompt, /cbm command
src/bridge.mjs # stdio JSON-RPC client, executable discovery, cbm_* tools
src/skills.mjs # bundled skill provider registration
skills/codebase-memory/SKILL.md # agent-facing skill
cordis.patch.yml # bundle patch that mounts the plugin
Every push to main runs the Publish workflow: it runs the test suite and
publishes the package to npm via Trusted Publishing (OIDC). If the current
version already exists, the workflow bumps the patch version and commits it back
with [skip publish] to avoid a loop; including [skip publish] anywhere in a
commit message skips the publish run entirely.
MIT © 2025 LoveDoLove
Codebase Memory MCP bridge plugin for DeepSeek Harness: knowledge-graph code search, snippets, architecture, call-path tracing, and auto-started graph visualization UI.
{ "mcpServers": { "lovedolove-dsh-codebase-memory-mcp": { "command": "npx", "args": [ "-y", "@lovedolove/[email protected]" ] } }}Инструменты
Списка инструментов пока нет. У удалённых серверов он появляется после первой проверки состояния.
FAQ
Скопируйте на этой странице конфигурацию stdio для Cursor, Claude Desktop, Claude Code или VS Code. Она использует установку издателя, а не прокси PlugMCP.
В этой записи не указано обязательных переменных окружения.
@lovedolove/dsh Codebase Memory работает локально через stdio.
Последняя проверка: 2026-11-15. Codebase Memory MCP bridge plugin for DeepSeek Harness: knowledge-graph code search, snippets, architecture, call-path tracing, and auto-started graph visualization UI.