OpenCode plugin for persistent memory using Supermemory.
Your agent remembers what you tell it - across sessions, across projects.
bunx opencode-supermemory@latest installThen authenticate with Supermemory:
bunx opencode-supermemory@latest loginCheck the connection any time:
bunx opencode-supermemory@latest statusOr let your agent do it - paste this into OpenCode:
Install opencode-supermemory by following https://raw.githubusercontent.com/supermemoryai/opencode-supermemory/main/README.md
Instructions for LLM Agents (click to expand)
If you're an LLM agent helping set up opencode-supermemory, follow these steps:
bunx opencode-supermemory@latest install --no-tuiThis will:
- Register the plugin in
~/.config/opencode/opencode.jsonc - Create the
/supermemory-initcommand
cat ~/.config/opencode/opencode.jsoncShould contain:
{
"plugin": ["opencode-supermemory"]
}If not, add it manually:
JSONC:
JSON:
{
"plugin": ["opencode-supermemory"]
}Run the browser authentication flow:
bunx opencode-supermemory@latest loginFor headless environments, users can still set SUPERMEMORY_API_KEY manually or create ~/.config/opencode/supermemory.jsonc:
{
"apiKey": "sm_...",
}Tell the user to restart OpenCode and run:
bunx opencode-supermemory@latest statusIf it is not connected, check:
- Is the user authenticated, or is
SUPERMEMORY_API_KEYset? - Is the plugin in
opencode.jsonc? - Check logs:
tail ~/.opencode-supermemory.log
Run /supermemory-init to have the agent explore and memorize the codebase.
On first message, the agent receives (invisible to user):
- Personal profile for the current project
- 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.
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.
Run /supermemory-init to explore and memorize your codebase structure, patterns, and conventions.
When context hits 80% capacity:
- Triggers OpenCode's summarization
- Injects project memories into summary context
- Saves session summary as a memory
This preserves conversation context across compaction events.
API key is <private>sk-abc123</private>
Content in <private> tags is never stored.
The supermemory tool is available to the agent:
| Mode | Args | Description |
|---|---|---|
add |
content, type?, scope? |
Store memory |
search |
query, scope? |
Search memories |
profile |
query? |
View user profile |
list |
scope?, limit? |
List memories |
forget |
memoryId, scope? |
Delete memory |
Scopes: user (personal memories for the current project), project (default)
Types: project-config, architecture, error-solution, preference, learned-pattern, conversation
OpenCode sends the same shared coding-agent entity context as Claude Code and
Codex. Personal and project memories are distinguished with sm_scope
metadata inside the shared repository container.
| Scope | Tag | Metadata |
|---|---|---|
| User | repo_{project-name}__{repository-hash} |
sm_scope: "personal" |
| Project | repo_{project-name}__{repository-hash} |
sm_scope: "project" |
The repository hash comes from the normalized Git origin remote, so Claude
Code, Codex, and OpenCode use the same container for the same repository.
Repositories with the same name but different remotes remain isolated. Without
an origin remote, OpenCode falls back to the repository's real filesystem path.
OpenCode also reads previous user_project_*, repo_<project-name>,
claudecode_project_*, codex_user_*, codex_project_*, opencode_user_*,
and opencode_project_* containers, so upgrading does not require a migration.
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,
// Legacy prefix retained when reading containers made by older versions
"containerTagPrefix": "opencode",
// Optional legacy personal container to keep reading
"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.
By default, new writes use:
- Repository tag:
repo_{project-name}__{hash(normalized-origin-remote)} - No origin remote:
repo_{project-name}__{hash(real-repository-path)}
Older {prefix}_user_* and {prefix}_project_* containers remain readable.
userContainerTag is treated as a legacy personal read. You can still override
the unified write container with projectContainerTag:
{
// Continue reading a personal container made by an older version
"userContainerTag": "my-team-workspace",
// Override the unified container used for new writes
"projectContainerTag": "my-awesome-project",
}This is useful when you want to:
- Preserve a legacy personal memory container
- 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
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"]
}bun install
bun run build
bun run typecheckLocal install:
{
"plugin": ["file:///path/to/opencode-supermemory"],
}tail -f ~/.opencode-supermemory.logMIT
{ "plugin": [ "opencode-supermemory", // ... other plugins ], }