---
title: OpenDots
description: How ClawMetry observes OpenDots — where its data lives, the environment variables that move it, what cost data is real, and the traps in its store.
eyebrow: Runtime reference
---

# OpenDots

OpenDots couples a local SQLite workspace with CopilotKit Intelligence to give a coding agent durable context across sessions. ClawMetry reads the local database in read-only mode; no code change to OpenDots is needed.

| | |
|---|---|
| **Runtime id** | `opendots` |
| **Category** | AI agent workspace |
| **Vendor** | CopilotKit |
| **Tier** | Paid tier (Starter and up) |
| **Upstream** | [github.com/CopilotKit/OpenDots](https://github.com/CopilotKit/OpenDots) |

## Where the data comes from

ClawMetry reads these locations, in order, and never writes to any of them.

- `<project>/data/opendots.sqlite (default; discovered under cwd and common project roots one level deep)`

**Format.** SQLite with `thread_bindings`, `dots`, `tasks`, `runs` and `events` tables. ClawMetry opens it read-only.

## Environment variables

| Variable | Effect |
|---|---|
| `CLAWMETRY_OPENDOTS_DB` | point at a specific database file |
| `DATABASE_PATH` | OpenDots' own path override |

See [Environment variables](/docs/config/environment/) for the variables that apply to every runtime.

## Tokens and cost

**Tokens.** None. Model IDs, token counts and costs are not persisted in the local database.

**Cost.** Unavailable. No cost data exists in the local store; `cost_status` is always `unavailable`.

:::note Why this section exists
Runtimes differ enormously in what they persist. Rather than showing a plausible number everywhere, ClawMetry reports what the store actually contains and marks the rest unknown. A cost of `$0.00` and a cost of *unknown* mean different things, and the dashboard keeps them different.
:::

## Sub-agents and delegation

Not exposed. No delegation structure is in the local database.

## Things worth knowing

- Intelligence service messages (outbound calls to CopilotKit's cloud) are not in the local database. Only dot-level activity is captured.
- Audit rows are dot-scoped, not session-scoped. ClawMetry groups by dot id to reconstruct a session view.
- Receipts arrive late: a run row may appear seconds after the dot row that spawned it. A freshly written session may look incomplete on the first read.
- Sample results in the `events` table store only the most recent value for each key; earlier values are overwritten in place.

## Verify it is being read

```bash
# Is this runtime detected, and would this install unlock it?
clawmetry runtimes --json | jq '.runtimes[] | select(.id=="opendots")'

# Why is it locked, if it is?
clawmetry runtimes --why opendots

# Any sessions ingested yet?
curl -s 'http://localhost:8900/api/local/sessions?limit=5' \
  | jq '.rows[] | {session_id, agent_id, events, cost_usd}'
```

If detection succeeds but no sessions appear, work through [Troubleshooting](/docs/ops/troubleshooting/) — the usual causes are a relocated data directory (set the override above), a session limit that is hiding older runs, or a runtime that simply has not written anything yet.