Integrate Long-Term Memory with Qoder via MCP

Prerequisites

• Qoder IDE or Qoder CLI is installed and functional.

• Python >= 3.10.

• uv package manager is installed (for on-demand startup via uvx, recommended).

• Obtained the AnalyticDB for PostgreSQLContext service API URL and token.

• Obtained the adbpgmem_mcp_selfhosted project source code adbpgmem-mcp-server-0.2.0.tar.gz.

  Extract the archive to a directory named adbpgmem_mcp_selfhosted. All paths in the following steps refer to this extracted directory.

Installation

MCP Server does not require pre-installation. uvx creates a temporary virtual environment from the local project path on demand and starts the service. You only need to ensure that the source path is accessible.

# Verify the source path is accessible
ls /path/to/adbpgmem_mcp_selfhosted/pyproject.toml
# The file info should be displayed, confirming the path is correct

Parameters

Configuration Source Priority

Method 1: Inject via the env Field During MCP Registration (Recommended)

Inject environment variables directly in the settings.json env field when registering the MCP Server. No additional configuration files are needed. See Register MCP Server.

Method 2: Manually Create a Configuration File

Create a ~/.qoder/adbpgmem.conf (global) or .qoder/adbpgmem.conf (project-level) configuration file:

# Global configuration (applies to all projects)
mkdir -p ~/.qoder
cat > ~/.qoder/adbpgmem.conf << 'EOF'
ADBPGMEM_API_URL="https://api-longmemory-cn-chengdu.opentrust.net"
ADBPGMEM_API_TOKEN="sk-your-token-here"
ADBPGMEM_USER_ID="your.username"
EOF

# Project-level configuration (applies to the current project only)
mkdir -p .qoder
cat > .qoder/adbpgmem.conf << 'EOF'
ADBPGMEM_API_URL="https://api-longmemory-cn-chengdu.opentrust.net"
ADBPGMEM_API_TOKEN="sk-your-token-here"
ADBPGMEM_USER_ID="your.username"
EOF

Configuration File Description

The configuration file uses the KEY="VALUE" format, one entry per line. If the configuration is missing, MCP Server logs a warning and continues to start, but API calls will fail.

Global Registration (Recommended)

Edit ~/.qoder/settings.json to apply the configuration to all projects:

{
  "mcpServers": {
    "adbpgmem": {
      "command": "uvx",
      "args": ["--from", "/path/to/adbpgmem_mcp_selfhosted", "adbpgmem-mcp-server"],
      "env": {
        "ADBPGMEM_API_URL": "https://api-longmemory-cn-chengdu.opentrust.net",
        "ADBPGMEM_API_TOKEN": "sk-your-token-here",
        "ADBPGMEM_USER_ID": "your.username"
      }
    }
  }
}

Field Description

Project-Level Registration

Create .qoder/settings.json in the project root directory. This configuration applies only to the current project and can be committed to version control for team sharing:

{
  "mcpServers": {
    "adbpgmem": {
      "command": "uvx",
      "args": ["--from", "/path/to/adbpgmem_mcp_selfhosted", "adbpgmem-mcp-server"],
      "env": {
        "ADBPGMEM_API_URL": "https://api-longmemory-cn-chengdu.opentrust.net",
        "ADBPGMEM_API_TOKEN": "sk-your-token-here",
        "ADBPGMEM_USER_ID": "your.username"
      }
    }
  }
}

Configure AGENTS.md Behavior Instructions (Optional)

Add the following content to AGENTS.md (Qoder's project-level AI behavior specification file, automatically loaded at each session start) in the project root directory. This guides Qoder AI to proactively use memory tools. Without this configuration, MCP tools still work, but Qoder AI will not proactively call them.

# MCP Servers

- **adbpgmem**: Persistent memory service. At the start of each session, use search_memories
  to search for relevant context. Use add_memory to save architectural decisions,
  debugging experience, and coding patterns. Use update_memory when memory content
  changes. Better to save too much than to miss something.

How to Use MCP Tools

Detailed Tool Usage

You: Remember: this project uses Go 1.22 + Gin framework, ORM uses GORM v2, database is PostgreSQL 15

Qoder AI calls add_memory → Server stores the memory

You: Remember these project decisions: we chose Kafka over RabbitMQ because we need
    message replay capability; chose PostgreSQL for JSONB support; frontend uses React + TypeScript

Qoder AI calls add_memory(infer=true) → Server extracts 3 independent atomic facts and stores them separately

You: Search for previous memories about database selection

Qoder AI calls search_memories → Returns semantically matched historical memories

Typical Usage Scenarios

You: Remember: this project uses Python 3.11 + FastAPI, database is PostgreSQL 15,
    cache uses Redis 7, message queue uses Kafka 3.5

Qoder AI calls add_memory → Saves tech stack info to the memory store

Session 1:
You: Remember: our APIs uniformly use RESTful style, pagination uses page/page_size,
    sorting uses sort=field:asc|desc, error responses include type/title/status/detail fields

Session 2 (days later, a new session):
You: Help me design a new user management API endpoint

Qoder AI: Based on the API design conventions saved earlier, I'll design the user management endpoint...
         [Automatically outputs the design following team conventions without needing repeated instructions]