claude-code-plugin-for-goodmem-a-memory-infra-for-agents — plugin
claude-code-plugin-for-goodmem-a-memory-infra-for-agents is a Claude Code plugin that connects to GoodMem memory infrastructure for managing embedders, rerankers, and memory spaces. It helps build RAG pipelines and research agents.
MIT ★ 10 updated 2026-10-20
/plugin marketplace add anthropics/claude-plugins-community
/plugin install claude-code-plugin-for-goodmem-a-memory-infra-for-agents@claude-communityWhat claude-code-plugin-for-goodmem-a-memory-infra-for-agents does
The plugin links Claude Code to a GoodMem server through an internal MCP server and dedicated skills. Agents can manage embedders and memory spaces directly in chat. It handles data ingestion from plain text or local files, and extracts content from documents via OCR. When configuring a model, passing model_identifier is enough. The built-in model registry fills in dimensions and endpoints on its own.
Connecting requires GOODMEM_BASE_URL and GOODMEM_API_KEY. You can export them in your shell or trigger goodmem_configure directly during the conversation. When connecting to an internal server using a self-signed certificate, declare the PEM file path in NODE_EXTRA_CA_CERTS before opening Claude Code.
SaaS providers such as OpenAI or Cohere require explicit credentials. Without an API key, the plugin halts with an error before contacting the server. It also bundles reference skills for the GoodMem Python and Java SDKs.
What it helps with
- set up memory spaces and configure embedders for RAG pipelines
- ingest research papers or text files into GoodMem storage
- extract document text using the integrated OCR tool
- retrieve historical incident resolutions for troubleshooting
- look up API signatures and code snippets for Python and Java SDKs
Example prompts for the agent
- Configure GoodMem with base URL https://my-server.com and API key gm_abc123.
- Create a new memory space for research papers and configure an embedding model.
- Show me how to query memories using the Java SDK via goodmem:java.
How to install claude-code-plugin-for-goodmem-a-memory-infra-for-agents
Plugin marketplace
/plugin marketplace add anthropics/claude-plugins-community/plugin install claude-code-plugin-for-goodmem-a-memory-infra-for-agents@claude-communityStep by step, with commands for each platform: Claude Code plugins: marketplaces, installing, and building your own
Questions about claude-code-plugin-for-goodmem-a-memory-infra-for-agents
How to install claude-code-plugin-for-goodmem-a-memory-infra-for-agents in Claude Code?
Run /plugin marketplace add anthropics/claude-plugins-community in Claude Code, then run /plugin install claude-code-plugin-for-goodmem-a-memory-infra-for-agents@claude-community.
What does this plugin do?
It connects Claude Code to a GoodMem server to manage embedders, rerankers, and memory spaces, ingest files or text, and access SDK reference documentation.
How do you configure a GoodMem server with a self-signed TLS certificate?
Set the NODE_EXTRA_CA_CERTS environment variable to the absolute path of the PEM certificate file before launching Claude Code, then restart the MCP connection.
Source text
SKILL.md · MIT
# GoodMem Claude Code Plugin A Claude Code plugin for [GoodMem](https://docs.goodmem.ai) — memory infrastructure for AI agents. With this plugin, you can operate GoodMem's memory infrastructure in plain English — create and manage embedders, rerankers, and LLMs, ingest memories from files, base64 strings, and plain text, and build knowledge-powered agents for RAG and Deep Research. Also includes Python and Java SDK references for writing GoodMem code. ## Use cases - **Deep Research agent** — Create an embedder and space, ingest research papers and documents, then ask complex questions that retrieve and synthesize information across your entire knowledge base. - **RAG pipeline** — Configure embedders, rerankers, and LLMs, store source materials as memories, and retrieve relevant context for answering questions. - **Self-improving support agent** — Ingest historical issue tickets and resolutions, search for solutions matching a new problem description, and iteratively store new discoveries and fixes as memories so future troubleshooting gets smarter over time. ## Installation **From the community marketplace** (preferred): ``` /plugin marketplace add anthropics/claude-plugins-community /plugin install goodmem@claude-community ``` **From this repo** (fallback — use only if the community marketplace is unavailable): ``` /plugin marketplace add PAIR-Systems-Inc/goodmem-claude-code-plugin /plugin install goodmem@goodmem-plugins ``` ## Configuration ### What you need to configure **GoodMem credentials** (required): - `GOODMEM_BASE_URL` — GoodMem server URL (e.g., `https://your-server.example.com`) - `GOODMEM_API_KEY` — GoodMem API key (starts with `gm_`) **TLS / self-signed certificates** (optional): If the server uses a private or self-signed CA, set `NODE_EXTRA_CA_CERTS` to the absolute path of its PEM certificate file before launching Claude Code. Restart the MCP connection after changing this setting. An expired or invalid server certificate must be fixed on the server. ### How to configure **1. In English from chat** (GoodMem credentials only; TLS variables cannot be set this way) Just tell Claude your server details — it will call `goodmem_configure` automatically: ``` > "Configure GoodMem with base URL https://my-server.com and API key gm_abc123" ``` Credentials persist for the session. You can reconfigure anytime to switch servers. **2. Shell export** (before launching Claude Code; works for all configurations) ```bash export GOODMEM_BASE_URL="https://your-server.example.com" export GOODMEM_API_KEY="gm_..." export NODE_EXTRA_CA_CERTS="/path/to/rootCA.pem" # only if needed for self-signed certs ``` ## Reloading after updates After installing a new version or pulling the latest changes, run `/reload-plugins` inside Claude Code to pick up the updates without restarting. If you are using the plugin from this repo, you can run `git pull` to get the latest changes. ## What's included | Component | Description | |-----------|-------------| | `skills/help/` | Setup guide, available skills overview, example workflows | | `skills/python/` | Python SDK reference — method signatures, parameters, examples | | `skills/java/` | Java SDK usage reference — setup, builders, streaming, pagination, examples | | `skills/mcp/` | MCP workflow guide; tool discovery provides current parameters | | `.mcp.json` | MCP server with auto-inference from the built-in model registry | ### Skills - **`goodmem:help`** — Overview of all skills, setup instructions, example workflows - **`goodmem:python`** — Python SDK reference for writing GoodMem code - **`goodmem:java`** — Java SDK reference for writing GoodMem code - **`goodmem:mcp`** — MCP tools reference for direct operations ### MCP tools The MCP server exposes all GoodMem API operations as tools: - **goodmem_configure** — set server credentials from chat - **goodmem_lookup_model** — look up model info from the built-in registry - **embedders** — create, list, get, update, delete embedding models - **llms** — create, list, get, update, delete LLM configurations - **rerankers** — create, list, get, update, delete reranker models - **spaces** — create, list, get, update, delete memory spaces - **memories** — create, list, get, update, delete, retrieve, batch operations - **ocr** — extract text from documents - **users, apikeys, system, admin** — manage users, API keys, server ### Auto-inference When creating embedders, LLMs, or rerankers, just provide `model_identifier` and the plugin auto-fills provider, endpoint, dimensions, and other fields from the built-in model registry. You only need to provide `display_name`, `model_identifier`, and credentials — everything else is inferred. Explicit values always override inferred defaults. ### Credential validation SaaS providers (OpenAI, Cohere, Voyage, Jina, and OpenAI-compatible endpoints for Anthropic, Google, Mistral) require API credentials. If you create an embedder, LLM, or reranker pointing at a known SaaS hostname without providing credentials, the plugin raises a clear error before the request reaches the server. Pass `credentials` (MCP), `api_key` (Python SDK), or the provider API key overload (Java SDK) to proceed. ## Links - [GoodMem Documentation](https://docs.goodmem.ai) - [Python SDK on PyPI](https://pypi.org/project/goodmem/) - [MCP Server on npm](https://www.npmjs.com/package/@pairsystems/goodmem-mcp)
© PAIR-Systems-Inc · MIT · PAIR-Systems-Inc/goodmem-claude-code-plugin