Skip to content
Back to search

clawmetry

See your agent think. Zero-config observability & governance for 30 AI agent runtimes: Claude Code, OpenAI Codex, Hermes, OpenClaw & 26 more. Live token costs, sessions, tool calls, crons.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today

Check it deeper on your machine

It has changed since its scan on 11 Nov 2026.

As a Claude Code plugin, install it and then ask Claude Code:

/plugin marketplace add keithmackay/mackayi/plugin install scanskillsafety@mackayiDo a deep scan of https://github.com/vivekchand/clawmetry with scanskillsafety

Or from a terminal, clone it and run it:

git clone https://github.com/keithmackay/scanskillsafety-skill ~/.claude/skills/scanskillsafetynode ~/.claude/skills/scanskillsafety/cli.cjs --deep 'https://github.com/vivekchand/clawmetry'

It checks the exact commit: every file, its packages against OSV (api.osv.dev), how it compares with this page's scan (findsafeskills.com receives the repository's address), and, if you agree, Claude reading every file with your own Claude subscription. It never runs the code.

Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/vivekchand--clawmetry/badge.svg)](/listings/vivekchand--clawmetry)

HTML

<a href="/listings/vivekchand--clawmetry"><img src="/api/listings/vivekchand--clawmetry/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

These steps run a script from the network that findsafeskills does not scan.

pip install clawmetry     # then: clawmetry

Or the one-liner: `curl -sSL https://raw.githubusercontent.com/vivekchand/clawmetry/main/install.sh | bash`

Needs Python 3.8+ on macOS, Linux or Windows, and at least one agent runtime on the same machine. Docker instructions: [docs/DOCKER.md](docs/DOCKER.md).

Or let the agent set it up for you. The [`agent-kill-switch`](skills/agent-kill-switch/SKILL.md) skill teaches Claude Code, Codex, Cursor, Gemini CLI, Copilot or OpenCode to install ClawMetry, report what the agents on the machine are doing and spending, stop one session on request, and hold risky tool calls for approval:

npx skills add vivekchand/clawmetry --skill agent-kill-switch

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 342 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

This file has changed since it was scanned, so this is not exactly the text that was scanned, and no line is marked. The findings are listed with the safety scan above.

  1. # ClawMetry
  2. [![PyPI version](https://img.shields.io/pypi/v/clawmetry?color=E5443A&label=version)](https://pypi.org/project/clawmetry/)
  3. [![PyPI Downloads](https://static.pepy.tech/badge/clawmetry)](https://clickpy.clickhouse.com/dashboard/clawmetry)
  4. [![GitHub stars](https://img.shields.io/github/stars/vivekchand/clawmetry?style=flat&color=E5443A)](https://github.com/vivekchand/clawmetry/stargazers)
  5. [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT)
  6. [![OpenSSF Scorecard](https://api.scorecard.dev/projects/github.com/vivekchand/clawmetry/badge)](https://scorecard.dev/viewer/?uri=github.com/vivekchand/clawmetry)
  7. [![Security policy](https://img.shields.io/badge/security-policy-informational)](SECURITY.md)
  8. [![Egress: documented](https://img.shields.io/badge/egress-documented-informational)](docs/EGRESS.md)
  9. **An agent can make a hundred tool calls without making progress.** ClawMetry
  10. reads the session files your coding agents already write, and puts the timeline,
  11. the tool calls and whatever token and cost data the runtime exposes into one
  12. view — so you can tell a long run that is working from one that is stuck.
  13. Works with **32 AI agent runtimes** — Claude Code, OpenAI Codex, Hermes, OpenClaw & 28 more. One dashboard for your whole agent fleet. ([the full list](SUPPORTED_RUNTIMES.txt), generated from the catalogue.)
  14. ## Get started
  15. **Download the desktop app.** Install it on the machine where your agents run, then open ClawMetry. No terminal commands required.
  16. [![Download for macOS](https://img.shields.io/badge/macOS-Download%20.dmg-E5443A?style=for-the-badge&logo=apple&logoColor=white)](https://github.com/vivekchand/clawmetry/releases/latest/download/ClawMetry-mac.dmg)
  17. [![Download for Windows](https://img.shields.io/badge/Windows-Download%20.exe-0078D4?style=for-the-badge)](https://github.com/vivekchand/clawmetry/releases/latest/download/ClawMetry-windows-setup.exe)
  18. [![Download for Linux](https://img.shields.io/badge/Linux-Download%20.AppImage-333333?style=for-the-badge&logo=linux&logoColor=white)](https://github.com/vivekchand/clawmetry/releases/latest/download/clawmetry-linux.AppImage)
  19. [Desktop setup steps](#desktop-app) · [Other install options](#install)
  20. **Free and paid plans:** OpenClaw, NVIDIA NemoClaw, Goose and Qwen Code are free, with no account required.
  21. Paid **Starter** adds the other supported runtimes and cloud fleet monitoring. **Pro** adds governance, evaluations and waste detection.
  22. **[Try Pro free for 7 days](https://clawmetry.com/connect)** (no credit card) · **[Compare plans and pricing](https://clawmetry.com/pricing)**
  23. **Prefer the terminal?** With Python installed, run:
  24. ```bash
  25. pip install clawmetry && clawmetry
  26. ```
  27. Open **http://localhost:8900**. ClawMetry finds supported agent runtimes automatically. Observation is the default. Controls are opt-in.
  28. > 🌐 **Read this in:** [English](README.md) · [简体中文](docs/i18n/zh-CN/README.md) · [日本語](docs/i18n/ja/README.md) · [한국어](docs/i18n/ko/README.md) · [Español](docs/i18n/es/README.md) · [Português (BR)](docs/i18n/pt-BR/README.md) · [Français](docs/i18n/fr/README.md) · [Deutsch](docs/i18n/de/README.md) · [हिन्दी](docs/i18n/hi/README.md) · [العربية](docs/i18n/ar/README.md) · [Русский](docs/i18n/ru/README.md) · [more →](docs/i18n/)
  29. ![ClawMetry dashboard: every AI agent runtime on one machine with 24h and lifetime cost per agent](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/hero.png)
  30. ## Before you install
  31. | | |
  32. |---|---|
  33. | **What it does** | Reads the session files and logs your agents already write. No SDK, no code change, no instrumentation in your app. |
  34. | **What you see** | Session timeline, tool-by-tool replay, token and cost breakdown, and trajectory signals (looping, repeated failures) — per runtime. |
  35. | **What's free** | `pip install clawmetry` reads **OpenClaw, NVIDIA NemoClaw, Goose and Qwen Code** with no account, no key and no network call. The other 28 — Claude Code, Codex, Cursor and the rest — are read by the closed-source `clawmetry-pro` companion, which arrives with the 7-day trial or a plan — see [docs/ENTITLEMENTS.md](docs/ENTITLEMENTS.md) for the exact split. |
  36. | **How to start** | `pip install clawmetry && clawmetry`, then open localhost:8900. No agents on this machine yet? `clawmetry --sample` opens on three labelled synthetic sessions. |
  37. | **What leaves your machine** | No session data, unless you run `clawmetry connect`. Two things do run by default, both opt-out and neither carrying session content: an anonymous install ping and a PyPI version check. Every destination is inventoried in [docs/EGRESS.md](docs/EGRESS.md), rebuilt from a wire capture rather than from reading comments. |
  38. Two limits worth knowing before you judge the output: runtimes expose very
  39. different data (some publish no cost at all — [the matrix](docs/compatibility.md)
  40. says which, per runtime), and observing an action is not the same as being able
  41. to block it ([which controls are real, per runtime](docs/APPROVALS.md)).
  42. ## Works with 32 agent runtimes
  43. **Free in the open source app:** 🦞 **[OpenClaw](https://clawmetry.com/runtimes/openclaw)** · 🟩 **[NVIDIA NemoClaw](https://clawmetry.com/nemoclaw)** · 🪿 **[Goose](https://clawmetry.com/runtimes/goose)** · ◈ **[Qwen Code](https://clawmetry.com/runtimes/qwen-code)**
  44. **On a paid plan:** ◆ **[Claude Code](https://clawmetry.com/runtimes/claude-code)** · **[Cursor](https://clawmetry.com/runtimes/cursor)** · 🐙 **[GitHub Copilot](https://clawmetry.com/runtimes/copilot)** · ⬡ **[OpenAI Codex](https://clawmetry.com/runtimes/codex)** · ♊ **[Gemini CLI](https://clawmetry.com/runtimes/gemini-cli)** · 💗 **[Lovable](https://clawmetry.com/runtimes/lovable)** · ⠕ **[Replit Agent](https://clawmetry.com/runtimes/replit)** · 🖇 **[Cline](https://clawmetry.com/runtimes/cline)** · 🙌 **[OpenHands](https://clawmetry.com/runtimes/openhands)** · 🧑‍💼 **[OpenWorker](https://clawmetry.com/runtimes/openworker)** · 🎭 **[Muse Code](https://clawmetry.com/runtimes/muse-code)** · 🏛️ **[OpenExecutive](https://clawmetry.com/runtimes/openexecutive)** · **[opencode](https://clawmetry.com/runtimes/opencode)** · **[Aider](https://clawmetry.com/runtimes/aider)** · 🔗 **[n8n](https://clawmetry.com/runtimes/n8n)** · 🅳 **[Devin](https://clawmetry.com/runtimes/devin)** · 🪐 **[Antigravity](https://clawmetry.com/runtimes/antigravity)** · **[Grok Build](https://clawmetry.com/runtimes/grok)** · 🤖 **[Grok Bot](https://clawmetry.com/runtimes/grok-bot)** · ⚡ **[Hermes](https://clawmetry.com/runtimes/hermes)** · **[Pi](https://clawmetry.com/runtimes/pi)** · **[Deep Agents](https://clawmetry.com/runtimes/deep-agents)** · 🌙 **[Kimi CLI](https://clawmetry.com/runtimes/kimi)** · 🐋 **[DeepSeek Harness](https://clawmetry.com/runtimes/deepseek-harness)** · 🦾 **[Exo](https://clawmetry.com/runtimes/exo)** · **[NanoClaw](https://clawmetry.com/runtimes/nanoclaw)** · **[PicoClaw](https://clawmetry.com/runtimes/picoclaw)** · **[QM](https://clawmetry.com/runtimes/qm)**
  45. Every runtime gets the same dashboard. Run several at once and the header
  46. switcher re-scopes every tab to one of them.
  47. Built your own agent on an SDK instead? The interceptor tracks its LLM calls
  48. too. See [docs/SDK_TRACKING.md](docs/SDK_TRACKING.md).
  49. Building with a humanoid robot? The free, open-source
  50. [Asimov 1 connector](docs/ASIMOV.md) imports Menlo SDK recordings and provides
  51. an optional observation collector and application tracing. This support is
  52. experimental. Physical hardware qualification is pending.
  53. [Jumper observation](docs/JUMPER.md) is also built into the same open-source
  54. package. Its application telemetry, task decisions and incidents appear in the
  55. shared Physical AI view. Mizu's separate scene and skills demo imports this
  56. built-in adapter; it is not another observer plugin to install.
  57. ## What you get
  58. - **Sessions & transcripts**: what each agent did, turn by turn, with replay
  59. - **Cost & tokens**: per runtime, model, session and day, with anomaly flags
  60. - **Flow**: live diagram of messages moving through channels, models and tools
  61. - **Brain**: the reasoning and tool-call event stream as it happens
  62. - **Context blowout**: window utilization sized per provider, compaction vs forced overflow, plus a per-runtime map of what we *can't* see ([how](docs/CONTEXT_BLOWOUT.md))
  63. - **Memory & skills**: the files and skills each runtime actually loaded
  64. - **Health & logs**: disk, memory, error rates, rate limits, live log stream
  65. - **Alerts**: budget caps, error spikes, agent-offline, routed to Slack, Discord, PagerDuty, Telegram, Email
  66. - **Approvals**: pause risky tool calls *before* they run and approve from your phone ([how](docs/APPROVALS.md))
  67. ## Context blowout, and what watching costs
  68. Two questions worth answering before you trust any agent-comparison tool.
  69. **How does it handle context-window blowout across runtimes?**
  70. A utilization percentage is only as honest as what it divides by. ClawMetry
  71. sizes the window per provider from [a table you can read and
  72. PR](clawmetry/context_windows.py), covering Anthropic, OpenAI, Google, xAI,
  73. DeepSeek, Kimi, Qwen, Mistral, Llama and GLM. It does not measure all 32
  74. runtimes with one vendor's ruler. That matters: a 300K GPT-5 turn scored
  75. against Anthropic's 200K reads ">100%, blown" when it is really at 75% of
  76. GPT-5's 400K. The same ruler hides a genuinely overflowed 130K DeepSeek turn
  77. as a comfortable 65%.
  78. Every window ships with its provenance: `model_table`, `explicit_marker`,
  79. `observed_floor`, or an honest `default` when we don't know the model. A
  80. gauge built on a guess never renders with the same authority as one built on
  81. a lookup.
  82. ClawMetry can only see compaction events on some runtimes. So
  83. `GET /api/context-coverage` reports, per runtime, whether a **zero means
  84. "ran clean" or "we're blind"**. A `0` that actually means blind says so.
  85. [Full detail](docs/CONTEXT_BLOWOUT.md)
  86. **What does the instrumentation cost?**
  87. | Path | Added to your agent | Default? |
  88. |---|---|---|
  89. | Session-file tailing (all 32 runtimes) | **0**. Separate process, no ClawMetry code in your agent | on |
  90. | HTTP interceptor (`CLAWMETRY_INTERCEPT=1`) | **+0.44 ms** per LLM call, or 0.009% of a 5s call | off |
  91. | Pre-tool hook gate (warm cache) | **+44 ms** per gated tool call, over a 36 ms interpreter floor | off |
  92. | Enforcement proxy | **+9.7 ms** per LLM call | off |
  93. Daemon host cost: **2,762 events/sec** ingest, **710 bytes/event** on disk
  94. (67.7 MB per 100k events), and **~12% of one core** sustained on a busy
  95. install. That last number is over our own stated 5-10% budget, so it is
  96. published as a bug to chase rather than left off the page.
  97. Measured on an Apple M2 Pro with `benchmarks/overhead.py`. The harness runs
  98. each condition in a separate process, alternates their order, and **refuses
  99. to print a number when the rounds disagree on its sign**. Run it on your own
  100. machine in a minute:
  101. ```bash
  102. pip install clawmetry && python -m benchmarks.overhead
  103. ```
  104. Every path is measured, including the hook gates and the enforcement proxy,
  105. and the harness runs on Linux, macOS and Windows in CI. Two results worth
  106. knowing: the proxy costs about seven times more on Windows than on Linux, and
  107. the daemon currently sustains about 12% of one core, over our own 5-10%
  108. budget. The raw JSON, the method, and what is still unmeasured are in
  109. [docs/OVERHEAD.md](docs/OVERHEAD.md).
  110. ## Pricing
  111. | Plan | What it covers | Price |
  112. |---|---|---|
  113. | **Free** | OpenClaw + NVIDIA NemoClaw + Goose + Qwen Code, full dashboard, local only | $0 |
  114. | **Starter** | Every other runtime above, fleet view, cloud sync | $9 per node / month |
  115. | **Pro** | Starter + control and evaluation: approvals, tool-risk policies, evals, anomaly detection, cost optimizer, OTel export, tamper-evident audit log | $19 per node / month |
  116. Annual plans, Enterprise and the current numbers live at
  117. **[clawmetry.com/pricing](https://clawmetry.com/pricing)**. Self-hosted license
  118. keys work without the cloud (`clawmetry license`). The exact free/paid split is
  119. in [docs/ENTITLEMENTS.md](docs/ENTITLEMENTS.md).
  120. ## Your data stays on your machine
  121. ClawMetry reads local session files and logs. **No session data leaves your box
  122. unless you run `clawmetry connect`** — no prompts, replies, tool arguments, file
  123. contents or log lines. When you do connect, the snapshot is end-to-end encrypted
  124. with a key that never leaves your machine, and decrypted in your browser. If a
  125. node has no key, the upload is skipped rather than sent in the clear, and no
  126. server response can turn that off.
  127. Two things do run by default before you connect, both opt-out and neither
  128. carrying session data: an anonymous install ping and a version check against
  129. PyPI. A default install also looks up your public IP once for a startup banner
  130. line. Every destination, what it carries and how to switch it off is listed in
  131. [docs/EGRESS.md](docs/EGRESS.md); self-hosted, repointed and air-gapped installs
  132. make no discretionary outbound calls at all.
  133. The decryption happens in your browser, in code we serve you. That used to be
  134. a promise; it is now something you can check. Every line that touches your key
  135. lives in one readable file, [`clawmetry/static/js/cm-e2e.js`](clawmetry/static/js/cm-e2e.js),
  136. which ships inside the wheel and is served verbatim, pinned with a Subresource
  137. Integrity hash. To confirm the browser runs what we published:
  138. ```bash
  139. curl -s https://app.clawmetry.com/static/js/cm-e2e.js -o served.js
  140. pip download --no-deps clawmetry==$(clawmetry --version | tr -d 'a-z ') -d /tmp/cm
  141. unzip -p /tmp/cm/clawmetry-*.whl clawmetry/static/js/cm-e2e.js > published.js
  142. diff served.js published.js && echo identical
  143. ```
  144. What that does not prove: we serve the page that loads the file, so we could
  145. serve a different page. Integrity hashes protect you from a compromised CDN,
  146. not from the vendor. What you gain is that any substitution has to be
  147. deliberate, visible in the page source, and different from an artifact on PyPI
  148. that anyone can fetch. Self-hosting or staying local-only removes the
  149. dependency entirely.
  150. ## Install
  151. ### Desktop app
  152. | Platform | Download | Open the app |
  153. |---|---|---|
  154. | **macOS** | [Download .dmg](https://github.com/vivekchand/clawmetry/releases/latest/download/ClawMetry-mac.dmg) | Open the disk image. Drag ClawMetry to Applications, then open it. |
  155. | **Windows** | [Download .exe installer](https://github.com/vivekchand/clawmetry/releases/latest/download/ClawMetry-windows-setup.exe) | Run the installer. Open ClawMetry from the Start menu. |
  156. | **Linux (x86_64)** | [Download .AppImage](https://github.com/vivekchand/clawmetry/releases/latest/download/clawmetry-linux.AppImage) | Allow the file to run as a program in its file properties, then open it. |
  157. The first launch needs an internet connection to install the current ClawMetry runtime.
  158. See [desktop requirements and troubleshooting](desktop/README.md) for Python requirements, signing status and platform details.
  159. ### Terminal
  160. ```bash
  161. pip install clawmetry
  162. clawmetry
  163. ```
  164. Or the one-liner: `curl -sSL https://raw.githubusercontent.com/vivekchand/clawmetry/main/install.sh | bash`
  165. Needs Python 3.8+ on macOS, Linux or Windows, and at least one agent runtime on
  166. the same machine. Docker instructions: [docs/DOCKER.md](docs/DOCKER.md).
  167. Or let the agent set it up for you. The [`agent-kill-switch`](skills/agent-kill-switch/SKILL.md)
  168. skill teaches Claude Code, Codex, Cursor, Gemini CLI, Copilot or OpenCode to
  169. install ClawMetry, report what the agents on the machine are doing and spending,
  170. stop one session on request, and hold risky tool calls for approval:
  171. ```bash
  172. npx skills add vivekchand/clawmetry --skill agent-kill-switch
  173. ```
  174. ## Docs
  175. | | |
  176. |---|---|
  177. | [Runtime compatibility](docs/compatibility.md) | What each adapter reads, and how to add a runtime |
  178. | [Context blowout](docs/CONTEXT_BLOWOUT.md) | Per-provider windows, compaction vs overflow, per-runtime coverage |
  179. | [Overhead](docs/OVERHEAD.md) | What instrumentation costs, measured, with the harness to reproduce it |
  180. | [Entitlements](docs/ENTITLEMENTS.md) | Free vs paid, tier matrix, license CLI |
  181. | [Approvals & policies](docs/APPROVALS.md) | Pre-execution gating, risk scoring, phone approvals |
  182. | [OpenTelemetry](docs/OPENTELEMETRY.md) | Export traces anywhere, ingest OTLP from anything |
  183. | [Bring your own agent](docs/BRING_YOUR_OWN_AGENT.md) | AWS AgentCore, Pydantic AI, LangChain end to end, with runnable examples |
  184. | [SDK tracking](docs/SDK_TRACKING.md) | Cost attribution for agents you built yourself |
  185. | [Chat channels](docs/CHANNELS.md) | The chat adapters shown in Flow |
  186. | [NemoClaw / OpenShell](docs/NEMOCLAW.md) | Sandboxed NVIDIA NemoClaw setups |
  187. | [Docker](docs/DOCKER.md) | Image, compose, volume mounts |
  188. | [Architecture](ARCHITECTURE.md) · [Development](docs/DEVELOPMENT.md) | How it works inside; running from source |
  189. | [Telemetry](docs/TELEMETRY.md) | The anonymous install and desktop-open pings, and how to turn them off |
  190. ## Screenshots
  191. Every number below is from one real machine, read-only, with nothing seeded.
  192. **It tells you when something is wrong, not just what happened.**
  193. Two anomaly banners at the top: spend running 7x the daily average, and a
  194. 4.2x cost spike. Below them, 324 of 667 recent sessions carrying a waste
  195. signal, itemised by cause.
  196. ![Overview: spending anomaly and cost spike banners over live agent work](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/overview.png)
  197. **It shows you where the money went, in every window.**
  198. $252.47 today, $513.15 this week, $1,312.92 this month, each with the tokens
  199. behind it and how much of it your subscription already covers. Below that,
  200. about $1,128/mo itemised as recoverable and $17,256/mo already saved by
  201. cache reuse.
  202. ![Cost: today, this week and this month, with an efficiency grade and itemised savings ideas](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/cost.png)
  203. **It draws how a message becomes an answer.**
  204. The live flow diagram: you, the channel it arrived on, the gateway, the model
  205. answering right now, and every tool it reached for. Nodes light up as work
  206. moves through them.
  207. ![Flow: live diagram from you through the gateway to the model and its tools](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/flow.png)
  208. **Every agent on the machine, in one table.**
  209. What it runs, what it costs in the last 24 hours and over its lifetime, when
  210. it was last seen, who owns it, and whether a subscription is covering the
  211. bill. 14 agents here, 3 sessions working, 13 quiet.
  212. ![Agents: every runtime on the machine with cost, owner, last seen and current work](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/agents.png)
  213. **It shows where a turn's time and money went, tool by tool.**
  214. One turn of a real session: 11 tools in 11.2 minutes for $1.16. Every Bash
  215. call and model call gets its own bar on the timeline, so the command that ran
  216. for 4.1 minutes and the one that ran for 226ms are told apart at a glance.
  217. ![Sessions: one agent turn on a timeline, every tool call with its own duration and the turn's cost](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/sessions.png)
  218. **It grades the work, not just the spend.**
  219. An A this week: 54 tasks came back clean, 2 rough ones cost $48.57, and the
  220. runs with too little activity to judge are left out of the grade instead of
  221. being counted as wins. Each rough run links to its trace.
  222. ![Quality: this week's report card with the rough runs and what they cost](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/quality.png)
  223. **It shows why the context window keeps filling up.**
  224. 715K of a 1M-token window on the latest turn, an 83.3% peak, 4 compactions
  225. that all fired proactively rather than on an overflow, and the utilisation of
  226. every turn behind it.
  227. ![Context usage: window utilisation per turn, compaction events and tokens reclaimed](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/context.png)
  228. **Detection runs without you configuring anything.**
  229. The built-in detectors are on from install: agent went quiet, telemetry feed
  230. stopped, cost spike, token burst, errors climbing, error spike, budget
  231. threshold, threat signature matched, security tool finding, security posture
  232. changed. Your own rules are optional on top.
  233. ![Alerts: built-in detectors plus optional custom rules](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/alerts.png)
  234. **Holding a risky call is opt-in, and ships off.**
  235. Recursive deletes, force pushes, sudo, secrets, package installs and outbound
  236. calls each get a rule you can turn on. Until you do, ClawMetry watches and
  237. changes nothing. Once one is on, matching calls wait here (or on your phone)
  238. for an approve or a deny.
  239. ![Approvals: protection rules for risky tool calls, all off until you enable them](https://raw.githubusercontent.com/vivekchand/clawmetry/main/screenshots/approvals.png)
  240. More, per runtime: [docs/RUNTIME_SCREENSHOTS.md](docs/RUNTIME_SCREENSHOTS.md).
  241. ## Recognition
  242. <a href="https://www.producthunt.com/products/clawmetry?embed=true&utm_source=badge-top-post-badge&utm_medium=badge&utm_campaign=badge-clawmetry-for-openclaw" target="_blank"><img src="https://api.producthunt.com/widgets/embed-image/v1/top-post-badge.svg?post_id=1081207&theme=light&period=daily&t=1771491508782" alt="ClawMetry - #5 Product of the Day on Product Hunt" width="250" height="54" /></a>
  243. ## Star History
  244. <a href="https://www.star-history.com/?repos=vivekchand%2Fclawmetry&type=date&legend=top-left">
  245. <picture>
  246. <source media="(prefers-color-scheme: dark)" srcset="https://api.star-history.com/image?repos=vivekchand/clawmetry&type=date&theme=dark&legend=top-left" />
  247. <source media="(prefers-color-scheme: light)" srcset="https://api.star-history.com/image?repos=vivekchand/clawmetry&type=date&legend=top-left" />
  248. <img alt="Star History Chart" src="https://api.star-history.com/image?repos=vivekchand/clawmetry&type=date&legend=top-left" />
  249. </picture>
  250. </a>
  251. ## License
  252. MIT · Built by [@vivekchand](https://github.com/vivekchand) · [clawmetry.com](https://clawmetry.com)
  253. <!-- osai-verify: f3ac716d40002c1ad6dd -->

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.