opencode-supermemory

OpenCode plugin for persistent memory using Supermemory.

Your agent remembers what you tell it - across sessions, across projects.

Installation

For Humans

bunx opencode-supermemory@latest install

Then authenticate with Supermemory:

bunx opencode-supermemory@latest login

Check the connection any time:

bunx opencode-supermemory@latest status

Or let your agent do it - paste this into OpenCode:

Install opencode-supermemory by following https://raw.githubusercontent.com/supermemoryai/opencode-supermemory/main/README.md

For LLM Agents

bunx opencode-supermemory@latest install --no-tui

cat ~/.config/opencode/opencode.jsonc

{
  "plugin": ["opencode-supermemory"]
}

{
  "plugin": [
    "opencode-supermemory",
    // ... other plugins
  ],
}

{
  "plugin": ["opencode-supermemory"]
}

bunx opencode-supermemory@latest login

{
  "apiKey": "sm_...",
}

bunx opencode-supermemory@latest status

Features

Context Injection

On first message, the agent receives (invisible to user):

• User profile (cross-project preferences)
• Project memories (all project knowledge)
• Relevant user memories (semantic search)

Example of what the agent sees:

[SUPERMEMORY]

User Profile:
- Prefers concise responses
- Expert in TypeScript

Project Knowledge:
- [100%] Uses Bun, not Node.js
- [100%] Build: bun run build

Relevant Memories:
- [82%] Build fails if .env.local missing

The agent uses this context automatically - no manual prompting needed.

Keyword Detection

Say "remember", "save this", "don't forget" etc. and the agent auto-saves to memory.

You: "Remember that this project uses bun"
Agent: [saves to project memory]

Add custom triggers via keywordPatterns config.

Codebase Indexing

Run /supermemory-init to explore and memorize your codebase structure, patterns, and conventions.

Preemptive Compaction

When context hits 80% capacity:

1. Triggers OpenCode's summarization
2. Injects project memories into summary context
3. Saves session summary as a memory

This preserves conversation context across compaction events.

Privacy

API key is <private>sk-abc123</private>

Content in <private> tags is never stored.

Tool Usage

The supermemory tool is available to the agent:

Scopes: user (cross-project), project (default)

Types: project-config, architecture, error-solution, preference, learned-pattern, conversation

Memory Scoping

Configuration

Create ~/.config/opencode/supermemory.jsonc:

{
  // API key (can also use SUPERMEMORY_API_KEY env var)
  "apiKey": "sm_...",

  // Supermemory API base URL (point at a self-hosted instance, e.g. http://localhost:8787)
  "baseUrl": "https://api.supermemory.ai",

  // Min similarity for memory retrieval (0-1)
  "similarityThreshold": 0.6,

  // Max memories injected per request
  "maxMemories": 5,

  // Max project memories listed
  "maxProjectMemories": 10,

  // Max profile facts injected
  "maxProfileItems": 5,

  // Include user profile in context
  "injectProfile": true,

  // Prefix for container tags (used when userContainerTag/projectContainerTag not set)
  "containerTagPrefix": "opencode",

  // Optional: Set exact user container tag (overrides auto-generated tag)
  "userContainerTag": "my-custom-user-tag",

  // Optional: Set exact project container tag (overrides auto-generated tag)
  "projectContainerTag": "my-project-tag",

  // Extra keyword patterns for memory detection (regex)
  "keywordPatterns": ["log\\s+this", "write\\s+down"],

  // Context usage ratio that triggers compaction (0-1)
  "compactionThreshold": 0.8,
}

All fields optional. Env var SUPERMEMORY_API_KEY takes precedence over config file.

Container Tag Selection

By default, container tags are auto-generated using containerTagPrefix plus a hash:

• User tag: {prefix}_user_{hash(git_email)}
• Project tag: {prefix}_project_{hash(directory)}

You can override this by specifying exact container tags:

{
  // Use a specific container tag for user memories
  "userContainerTag": "my-team-workspace",

  // Use a specific container tag for project memories
  "projectContainerTag": "my-awesome-project",
}

This is useful when you want to:

• Share memories across team members (same userContainerTag)
• Sync memories between different machines for the same project
• Organize memories using your own naming scheme
• Integrate with existing Supermemory container tags from other tools

Usage with Oh My OpenCode

If you're using Oh My OpenCode, disable its built-in auto-compact hook to let supermemory handle context compaction:

Add to ~/.config/opencode/oh-my-opencode.json:

{
  "disabled_hooks": ["anthropic-context-window-limit-recovery"]
}

Development

bun install
bun run build
bun run typecheck

Local install:

{
  "plugin": ["file:///path/to/opencode-supermemory"],
}

Logs

tail -f ~/.opencode-supermemory.log

License

MIT