How do I monitor Block Goose sessions?

Install ClawMetry (pip install clawmetry). The daemon auto-detects ~/.local/share/goose/sessions and renders every session, tool call, model, and token in one dashboard at http://localhost:8900.

Does ClawMetry track which Goose extension served each turn?

Yes. Goose's session log records the extension (MCP server) that provided each tool call; ClawMetry surfaces per-extension usage on the Sub-agents tab so you can see extension spread per session and catch an extension that is silently overactive.

Can I see Goose cost per session across all my machines?

Yes with ClawMetry Cloud. Sync is E2E encrypted; the cloud dashboard shows per-node, per-runtime, per-session Goose spend rolled up across every machine you run Goose on.

Does ClawMetry work with Goose's provider + model routing?

Yes. Goose's model routing (per-provider fallback, per-turn override) is visible on the Brain tab; per-model cost rolls up on the Cost tab.

How much does ClawMetry cost for Goose?

Free forever on your own machine. Cloud starts at $9/node/mo (Starter) or $19/node/mo (Pro).

Observability and cost tracking for Block Goose

See every Goose session, tool call, and token as it happens. ClawMetry reads Goose's session store under ~/.local/share/goose/sessions and renders sessions, tools, tokens, and model routing in one dashboard. No SDK.

Supported runtime Reads Session files written by Block's Goose Read-only Free trial

What ClawMetry sees on Goose

ClawMetry's Goose adapter opens the store Goose already writes (~/.local/share/goose/sessions) in read-only mode and surfaces:

Sessions: started at, last activity, model, message count, per-session token totals.
Live activity: every tool call as it lands, with args and result. Turn-level tokens split.
Per-turn tokens: Per-turn usage counters written by Goose (input, output, cache-read, cache-write).
Cost per session, per model, per hour: cache-aware where the model supports caching.
Per-extension attribution (Goose's extensions system: which MCP server + tool served each turn): the Goose-specific signal ClawMetry surfaces that generic dashboards do not.
Approval and kill switch: watch for token-maxing signals (message repeat, no-progress heuristic, cost-per-hour threshold), pause or kill before the next call.

Source: the adapter is at clawmetry_pro/adapters/goose.py. Every parse rule and every SQL SELECT is in that file.

Enable it in about two minutes

1. Install ClawMetry

Pick the line that matches your OS. Windows users: paste the install.cmd line, not the shell one.

# macOS or Linux
curl -fsSL https://clawmetry.com/install.sh | bash

# Windows (cmd.exe or PowerShell)
curl -fsSL https://clawmetry.com/install.cmd -o install.cmd && install.cmd

2. Run the daemon

clawmetry

The daemon auto-detects Goose's local store and starts reading immediately.

3. Open the dashboard

Open http://localhost:8900. Switch the runtime chip to Goose.

Start the free trial →

The one thing this catches that Goose alone does not

Left alone, Goose's failure mode is an extension left enabled that quietly reasons its way through unrelated turns. ClawMetry watches for it: cost per hour on the Goose runtime, message-repeat detection, and per-session token-growth curves. Set a threshold, and the approval queue pauses before the next call. Kill switch (Pro) stops it.

All 19 runtimes ClawMetry observes

Goose is one of 19. Free on every plan: OpenClaw, NVIDIA NemoClaw. Paid on Cloud Pro or Self-Hosted Pro: the rest.

OpenClaw NVIDIA NemoClaw NanoClaw PicoClaw Claude Code Codex Cursor Aider Goose (this page) opencode Qwen Code Hermes Pi Deep Agents n8n Antigravity GitHub Copilot Grok QM