Integration

Long-term memory for Cline

Cline is the open-source coding agent that ships as a VS Code extension and a CLI in one. Both reset every session. Add Engram to `cline_mcp_settings.json` with `autoApprove` set for the read-only memory tools, and Cline can store and recall without approval prompts breaking your flow.

Get an API key (free) All integrations →

Install

Three steps: sign up for an Engram API key, paste a BYOK LLM-provider key on /models, then drop the snippet below into Cline.

Three steps to memory in your agent

Sign up. Free, no card. You'll land on a Getting Started page that walks the next two steps.
Add your LLM key. Engram is BYOK. Paste an OpenAI / Anthropic / Groq / Together / Fireworks key and we'll route every extraction and query call through your provider. You pay your provider directly. We never see your inference.
Paste the snippet below into your agent and restart it. Use Authorization: Bearer <api-key>with the API key from your portal.

engram-cline-plugin: Cline integration (CLI + VS Code)

Six Engram tools wired into Cline's MCP surface. Cline namespaces the tools as engram__store_memory, engram__query_memory, etc. Source: github.com/lumetra-io/engram-cline-plugin.

Edit ~/.cline/data/settings/cline_mcp_settings.json (CLI) or paste into the Cline VS Code sidebar → MCP Servers → Configure:

cline_mcp_settings.json

{
  "mcpServers": {
    "engram": {
      "url": "https://mcp.lumetra.io/mcp/sse",
      "headers": {
        "Authorization": "Bearer <api-key>"
      },
      "disabled": false,
      "autoApprove": [
        "store_memory",
        "query_memory",
        "list_memories",
        "list_buckets"
      ]
    }
  }
}

Verify with cline config mcp; should print engram [sse].
Note: Cline doesn't support ${env:VAR} interpolation in headers, so the key lives in the JSON file as plaintext. The file isn't in any workspace and isn't committed to git, but treat it as a secret.

What you can do once memory's wired in

Recall the file-by-file plan from the last refactor and pick up where you stopped
Remember the codebase invariants you've taught Cline (e.g., 'always update both X and Y when changing Z')
Pull up reproductions and root causes for bugs we've already solved
Skip re-deriving the test-data fixtures every session by storing them once and recalling them

FAQ

How does Cline expose Engram tools to the model?

Cline wraps every MCP tool call in an XML `<use_mcp_tool>` block with separate `<server_name>engram</server_name>` and `<tool_name>store_memory</tool_name>` fields. You'll see those server/tool pairs in Cline's tool-approval UI. Internal provider-side tool IDs use a generated serverKey with Cline's own separator; that's an implementation detail, not something you configure.

What does `autoApprove` do?

Whitelists the four non-destructive tools (`store_memory`, `query_memory`, `list_memories`, `list_buckets`) so Cline doesn't prompt on every call. The destructive `delete_memory` and `clear_memories` still ask.

Does this work for both the CLI and the VS Code extension?

Yes, same JSON shape. The CLI reads `~/.cline/data/settings/cline_mcp_settings.json`; the VS Code extension reads the equivalent path under `globalStorage/saoudrizwan.claude-dev`. Configure one or both.

Related integrations

Ship durable memory in Cline today

Free tier: 10K memories and 50K retrievals per month. No credit card. Same Engram backend powers all 41 integrations, so memories you write from one client are immediately queryable from the rest.

Start free See pricing