How do I monitor OpenExecutive?

Install ClawMetry (pip install clawmetry) and run it. The daemon finds an OpenExecutive clone in your home folder or a common projects folder on its own and renders every conversation, specialist consult, tool call, token count and scheduled send in one dashboard at http://localhost:8900. No code change to OpenExecutive, no SDK and no proxy.

Does ClawMetry see OpenExecutive conversations that started in Slack or email?

Yes. Every conversation shows up, not only web chat: ones that started in Slack or email appear beside the rest. Each time the Executive consults a specialist agent, such as the CFO, the General Counsel or the CMO, that consult is a step in the session with the question it asked and the answer it got back.

Can I see what OpenExecutive is about to send on its own?

Yes. OpenExecutive has a scheduler that sends messages later on its own. ClawMetry lists every outbound send the Executive queued, by email, Slack, Telegram or Discord, with its time and its status: pending, delivered or failed. A pending send is visible before it goes out.

How accurate is the cost figure for OpenExecutive?

It is a floor, not the full bill. When the Executive's model calls go through OpenRouter, ClawMetry shows the real OpenRouter charge; otherwise it derives the cost from published model prices. OpenExecutive records usage for the Executive's own model calls only. The specialist agents' calls are not recorded, so their spend is not in the figure.

Is ClawMetry safe to point at a running OpenExecutive, and can it pause a conversation?

It is safe to point at a running install: ClawMetry only reads what OpenExecutive already stored, never writes to it and never sends a message on its behalf. It does not offer Pause, Stop or Kill for OpenExecutive, because one API server process serves every conversation and the scheduler. It says so rather than showing a control that would not do what it claims.

OpenExecutive × ClawMetry

Observability and cost tracking for OpenExecutive

See every OpenExecutive conversation, specialist consult, tool call and scheduled send. ClawMetry finds OpenExecutive on your machine and renders conversations from every channel, token counts and cost per session in one dashboard, next to every other agent you run.

Supported runtime Beta adapter Auto-detected in a local clone Read-only Free trial Official OpenExecutive repo ↗

📦 833k+ installs 🌍 129+ countries ⭐ 415 on GitHub 🏆 #5 Product Hunt

What ClawMetry sees on OpenExecutive

OpenExecutive is the open source AI executive team from Sente Labs, Apache-2.0 licensed. One Executive persona takes your question and routes it to specialist agents: a CFO, a CSO, a COO, a General Counsel, a CMO, a CPO, a CHRO and board communications. It answers over web chat, Slack, email, Telegram, Discord and Google Chat, and a scheduler lets it send messages later on its own. Once ClawMetry is running, that activity shows up as:

Every conversation, from every channel: not only web chat. A conversation that started in Slack or arrived by email is a session like any other, in the same list.
Specialist consults as steps: each time the Executive asks a specialist, the consult appears in the session as its own step, with the question it asked and the answer it got back.
Tool calls, with failures flagged: every tool call in the order it happened, and the ones that failed are marked so they do not hide inside a long session.
Tokens and cost per session: the real OpenRouter charge when the Executive's calls go through OpenRouter, otherwise a cost derived from published model prices.
Every scheduled send, before it goes out: each outbound message the Executive queued by email, Slack, Telegram or Discord, with its time and its status: pending, delivered or failed. A pending send is on screen before anyone receives it.

Two limits, stated plainly. The cost figure covers the Executive's own model calls. OpenExecutive does not record the specialist agents' calls, so the figure is a floor, not the full bill. And ClawMetry does not offer Pause, Stop or Kill for OpenExecutive: one API server process serves every conversation and the scheduler, so there is no single conversation to hold.

Read-only by design: ClawMetry reads the store OpenExecutive already keeps. No code change to OpenExecutive, no SDK and no proxy. It never writes to OpenExecutive and never sends a message on its behalf. The OpenExecutive adapter is in beta.

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

Or skip the terminal entirely and grab the desktop app: macOS (.dmg) · Windows (.exe) · Linux (.AppImage). It installs and updates ClawMetry for you.

2. Run the daemon

clawmetry

If you run OpenExecutive from a clone in your home folder, or in a projects folder named projects, src, code, dev, repos or workspace, the daemon finds it on its own. Nothing to add to OpenExecutive.

Running OpenExecutive in Docker? A named volume is not visible from the host. Bind-mount the container's /data directory to a folder on the host, or point ClawMetry at the database before you start it:

export CLAWMETRY_OPENEXECUTIVE_DB=/path/to/episodic_memory.db
clawmetry

3. Open the dashboard

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

Start the free trial →

What this shows that OpenExecutive alone does not

OpenExecutive answers the question in front of you. What one conversation cannot show is everything around it: the conversations that came in through Slack and email, what the Executive asked each specialist and what it heard back, which tool calls failed, and what it has queued to send later. ClawMetry keeps that view in one place, lists pending sends before they go out, and puts OpenExecutive's cost next to the other agents on the same machine, such as Claude Code and Codex.

Starred by the people building AI

Engineers and founders from OpenAI, Google, PostHog and more have starred ClawMetry on GitHub. One of them helped build it.

Peter SteinbergerOpenAI

Creator of OpenClaw and founder of PSPDFKit, now at OpenAI.

Matt Van HornJune · Weber

Co-founder of June, the smart oven acquired by Weber. ClawMetry contributor.

Alex KrentselGoogle

Systems research engineer at Google and PhD student at UC Berkeley.

Manoel Aranda NetoPostHog

Software engineer at PostHog, previously mobile SDKs at Sentry.

Kody KendallLlamaPress AI

Co-founder of LlamaPress AI, an open-source AI web-app builder.

All 32 runtimes ClawMetry observes

OpenExecutive is one of them; the other 31 are listed below. Free on every plan: OpenClaw, NVIDIA NemoClaw, Goose. Paid on Cloud Pro or Self-Hosted Pro: the rest.

Every runtime on one page, with what each one can and cannot show →