Cloud Sync

By default, l6e runs fully local — session data stays in SQLite on your machine and run logs are written to ~/.l6e/runs.jsonl. Cloud sync sends session metadata to the l6e dashboard so you get run history, usage charts, and calibration.

Getting an API key

1. Sign up at app.l6e.ai
2. Go to Settings → API Keys
3. Create a key — it starts with sk-l6e-

Enabling cloud sync

Set two environment variables in your MCP server config:

{
  "env": {
    "L6E_API_KEY": "sk-l6e-...",
    "L6E_CLOUD_SYNC": "1"
  }
}

See the setup guide for your client (Cursor, Claude Code, Windsurf, OpenClaw) for the full config location.

After setting these, restart your client. Cloud sync activates on the next l6e_run_end call.

What syncs

Data	Synced	Notes
Session ID and timestamps	Yes	When the session started and ended
Model and client identifiers	Yes	Which model and MCP client were used
Per-call token estimates	Yes	The estimates the agent provided at each checkpoint
Per-call gate decisions	Yes	allow, reroute, halt for each call
Total estimated cost	Yes	Accumulated spend for the session
Task summaries	Configurable	Controlled by L6E_SEND_TASK_SUMMARIES (default: true). Set to false to omit.
Prompts and completions	Never	l6e never sees or stores prompt content
Source code	Never	l6e has no access to your codebase

What you get

• Run history — every session with cost, call count, gate decisions, and exactness state
• Usage charts — spend over time, broken down by model and client
• Calibration — import billing data and l6e computes your personal calibration factor. See Calibration for details.
• Reconciliation — match l6e sessions against provider billing to see how estimates compare to reality

Cloud sync is also the mechanism that enables automatic calibration. When cloud sync is on, l6e_authorize_call gate decisions use your personal calibration factor from l6e.ai — computed from your billing imports. See Calibration for the calibration paths and l6e.ai Integration for the full setup guide.

How sync works

Session reports sync asynchronously via a local outbox. Reports are queued at l6e_run_end and drained in a background thread on the next l6e_run_start. This means cloud sync never blocks your agent's workflow — data appears in the dashboard after the next session starts. If the network is down, reports accumulate locally and sync when connectivity returns.

Disabling cloud sync

Remove L6E_CLOUD_SYNC (or set it to false) and restart your client. Sessions continue to be stored locally. Previously synced data remains in the dashboard — delete it from Settings → Data if needed.

Privacy

l6e syncs session metadata and cost data. It never receives, stores, or transmits prompt content, completions, or source code. The MCP server has no access to the LLM provider's request or response payloads — it only sees the token estimates and tool parameters that the agent passes to l6e_authorize_call.

Task summaries (the 5-10 word labels passed to l6e_run_start and l6e_run_end) are synced by default. Set L6E_SEND_TASK_SUMMARIES=false to omit them.