clawtel-setupUse when setting up clawtel to report token usage from a project that calls the Anthropic API (SDK, Claude Code, or any tapes-wrapped agent) to the claw.tech...
Install via ClawdBot CLI:
clawdbot install bdougie/clawtel-setupGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Sends data to undocumented external endpoint (potential exfiltration)
send → https://ingest.claw.tech/v1/heartbeat`.Potentially destructive shell commands in tool definitions
curl -fsSL https://download.tapes.dev/install | bashCalls external URL not in known-safe list
https://github.com/bdougie/clawtelUses known external API (expected, informational)
api.anthropic.comGenerated Oct 7, 2026
A developer running multiple OpenClaw-based agents (clawchief, staffchief, openclaw-in-a-box) needs each agent's Anthropic token usage reported to the claw.tech leaderboard. They hit the silent bypass bug where ANTHROPIC_BASE_URL is ignored and must patch ~/.openclaw/openclaw.json to point the provider baseUrl at the local tapes proxy, then verify via ss and sqlite3 that heartbeats carry real model and token counts.
An individual developer who uses Claude Code heavily across many side projects wants their cumulative token consumption to appear on claw.tech so they can compare against peers. They install tapes, launch Claude Code with ANTHROPIC_BASE_URL=http://localhost:8080, install the clawtel binary via the verified checksum path, and set CLAW_INGEST_KEY plus CLAW_ID so heartbeats start flowing within minutes.
A small team shipping a Node or Python product built directly on @anthropic-ai/sdk or the anthropic package wants to publicly showcase how many tokens their app processes. They wrap the client with the tapes proxy by exporting ANTHROPIC_BASE_URL or passing baseURL to the constructor, run clawtel as a persistent systemd or launchd service, and monitor claw.tech for heartbeat continuity after each deploy.
An operator notices the claw.tech dashboard has gone flat for a running agent despite healthy traffic. They follow the skill's diagnostic path: inspect the tapes SQLite nodes table for recent created_at rows, confirm the gateway process is peering to 127.0.0.1:8080 instead of Anthropic edge IPs, and check whether the ANTHROPIC_BASE_URL env var is set in the process that actually spawns the Anthropic client rather than just the interactive shell.
An engineering org building a staffchief-style agent wants monthly internal rankings of which teams and agents burn the most tokens. They standardize on clawtel plus tapes across every agent, use CLAW_ID to distinguish agents on the leaderboard, and keep tapes installed as a proxy so no prompt or response content ever leaves the machine while aggregate counts still reach claw.tech.
clawtel itself is a free single Go binary distributed via GitHub releases, while claw.tech is a hosted service that aggregates heartbeats into a public leaderboard. Value accrues from network effects and visibility rather than from charging for the binary.
clawtel acts as the connectivity layer between tapes (proxy/SQLite) and claw.tech, making it a natural adoption hook for teams already using openclaw, staffchief, or Claude Code. It reduces friction to publicly demonstrate AI usage, which drives word-of-mouth among builder communities.
Because clawtel only reads aggregate token counts and never prompt or response content, it fits privacy-conscious teams that still want operational visibility. Enterprises can adopt it as a low-risk observability layer for any Anthropic-calling workload.
💬 Integration Tip
For OpenClaw workloads, remember that ANTHROPIC_BASE_URL is ignored — you must set models.providers.anthropic.baseUrl in ~/.openclaw/openclaw.json and restart the gateway, then verify with `ss -tnp` that the peer is 127.0.0.1:8080. Always confirm tapes is capturing rows (`sqlite3 ~/.tapes/tapes.sqlite 'SELECT count(*), max(created_at) FROM nodes;'`) before troubleshooting clawtel itself.
Scored Oct 7, 2026
Audited Apr 21, 2026 · audit v1.0
Control desktop applications on Windows — launch, close, focus, resize, move windows, simulate keyboard/mouse input, manage processes, control VSCode, read clipboard, and capture screen info. Use when the user wants to interact with any running program, switch windows, type text, press shortcuts, open files in VSCode, manage running processes, or get system display information.
Conduct rigorous, adversarial code reviews with zero tolerance for mediocrity. Use when users ask to "critically review" my code or a PR, "critique my code", "find issues in my code", or "what's wrong with this code". Identifies security holes, lazy patterns, edge case failures, and bad practices across Python, R, JavaScript/TypeScript, SQL, and front-end code. Scrutinizes error handling, type safety, performance, accessibility, and code quality. Provides structured feedback with severity tiers (Blocking, Required, Suggestions) and specific, actionable recommendations.
Pragmatic coding standards for writing clean, maintainable code — naming, functions, structure, anti-patterns, and pre-edit safety checks. Use when writing new code, refactoring existing code, reviewing code quality, or establishing coding standards.
Claude Code integration for OpenClaw. This skill provides interfaces to: - Query Claude Code documentation from https://code.claude.com/docs - Manage subagents and coding tasks - Execute AI-assisted coding workflows - Access best practices and common workflows Use this skill when users want to: - Get help with coding tasks - Query Claude Code documentation - Manage AI-assisted development workflows - Execute complex programming tasks
Plan, draft, version, and refine written content with enforced versioning and quality audits.
Use when writing tests, creating test strategies, or building automation frameworks. Invoke for unit tests, integration tests, E2E, coverage analysis, performance testing, security testing.