Clawhub Rate Limit Exceeded: What It Means & How to Fix It

Quick Fix

Not sure which error you have? Start here:

1. If the error appears while installing or updating skills → you hit the ClawHub registry limit. Wait 60–120 seconds and retry.
2. If the error appears while chatting or running an AI agent → you hit the Anthropic/Claude API limit. Wait for the window to reset (see the Claude/Anthropic section below).
3. Run clawhub login if you haven't authenticated — authenticated users get a higher limit in both cases.
4. Update your CLI — older versions are significantly more "chatty" and hit limits faster:

   npm i -g clawhub@latest
   # or
   npx clawhub@latest install <skill>

Two Different Rate Limits

When you see "rate limit exceeded" in a ClawHub or OpenClaw context, there are two completely different things that can trigger it. Confusing them is the most common reason fixes don't work.

ClawHub Registry Rate Limit

This is the limit most people associate with clawhub install and clawhub update. The ClawHub.ai registry throttles requests per IP (and per authenticated account) to prevent abuse.

What the error looks like

Error: rate limit exceeded

clawhub install failed: HTTP 429 Too Many Requests

Common Causes

1. clawhub update --all

Running update --all fires one API request per installed skill simultaneously. If you have 20+ skills, this easily exceeds the registry limit in a single command.

2. Rapid back-to-back installs

Installing many skills in quick succession in a script:

clawhub install skill-a
clawhub install skill-b
clawhub install skill-c
# ...hits rate limit around here

3. Multiple parallel processes

Running clawhub commands across multiple terminal windows, or in CI jobs that fan out in parallel, stacks up requests from the same IP address.

4. Shared IP environments (VPS, corporate network, VPN)

If you're on a cloud server, corporate network, or VPN, many users may share the same outbound IP. The registry sees all their requests combined. This is a known issue — even a handful of requests from a shared IP can trigger the limit. Try:

• Disabling your VPN temporarily to test from your real IP
• Switching to a different network (mobile hotspot)

5. Old CLI version making extra requests

Older versions of the ClawHub CLI are more "chatty" — they make additional metadata requests that newer versions batch or skip. Update first:

npm i -g clawhub@latest

6. Network retries on a flaky connection

A slow or unstable connection can cause the CLI to retry in a loop, burning through your limit faster than expected. The Retry-After header handling was fixed in recent CLI versions — another reason to update.

Fixes

Fix 1: Wait and retry

The simplest fix. The rate limit window resets within 60–120 seconds in most cases.

# After waiting ~90 seconds:
clawhub install my-skill

Fix 2: Update your CLI

This single step fixes many rate limit issues. Older versions make redundant requests that newer versions have optimized away:

npm i -g clawhub@latest
# verify
clawhub --version

Fix 3: Log in to your Clawhub account

Authenticated users get a higher rate limit threshold than anonymous requests:

clawhub login
clawhub whoami  # confirm you're authenticated

Fix 4: Update skills one at a time with a delay

Instead of update --all, update individually with a pause between each:

clawhub update skill-a
clawhub update skill-b

Or add a short delay between installs in a shell script:

for skill in skill-a skill-b skill-c; do
  clawhub install "$skill"
  sleep 3
done

Fix 5: Check what's already installed

Before hitting the registry, check your local skills — you may already have it:

clawhub list

Fix 6: Switch network or disable VPN

If you're on a VPS or shared corporate network, your IP is already heavily rate-limited by shared usage. Try running from a different connection:

# Test from a different network, then:
clawhub install my-skill

Fix 7: Use a custom registry

If you run your own ClawHub registry instance, point the CLI at it to bypass the public rate limits entirely:

clawhub install my-skill --registry https://your-registry.example.com
# or set it persistently:
export CLAWHUB_REGISTRY=https://your-registry.example.com

Claude / Anthropic API Rate Limit

If you're using OpenClaw as an AI agent (chatting, running automations, processing files), the "rate limit exceeded" may be coming from Anthropic's API, not from ClawHub's skill registry. These are two completely separate systems.

What the error looks like

HTTP 429: rate_limit_error

Anthropic quota/rate limit is exhausted for the current window

Extra usage is required for long context requests

Why this happens

• Subscription users (setup-token): You've hit the usage ceiling for your current billing window. Claude subscriptions have tiered token budgets that reset on a rolling basis.
• API key users: You've hit your Anthropic Console rate limit tier. Free and lower tiers have strict requests-per-minute and tokens-per-day limits.
• Multiple simultaneous sessions: Running multiple OpenClaw windows or agents at the same time means they all draw from the same API key's rate limit pool.
• Long context requests: Processing very long documents or conversations consumes tokens faster and can trigger a specific long-context quota limit.

Fixes

Fix 1: Wait for the window to reset

Anthropic rate limit windows typically reset within a few minutes. For subscription users, the window is usually 1–5 minutes.

Fix 2: Check your usage in the Anthropic Console

If you're using an API key, log in to console.anthropic.com and check your current usage vs. limits. You can request a limit increase there.

Fix 3: Upgrade your plan

If you're consistently hitting limits, upgrading to a higher Anthropic tier or enabling Extra Usage on your subscription gives you a larger token budget.

Fix 4: Reduce simultaneous sessions

If you have multiple OpenClaw instances or agent pipelines running, pause them and consolidate. All sessions using the same API key share the same rate limit pool.

Fix 5: Configure a fallback model

In your OpenClaw settings, configure a fallback to a model from a different provider (not just a different Claude model). Falling back from Claude Sonnet to Claude Opus won't help — they share the same Anthropic rate limit. A fallback to a different provider keeps OpenClaw functional during Anthropic rate limit events.

How to Avoid Rate Limits (Both Types)

Registry rate limits:

• Keep your CLI updated: npm i -g clawhub@latest
• Always run clawhub login at the start of any session or CI pipeline
• Use clawhub list before installing to skip already-installed skills
• Add sleep 2–5 between installs in automation scripts
• Don't run clawhub in parallel across multiple processes on the same machine
• Use clawhub install skill-a skill-b skill-c (single command, multiple skills) instead of separate calls when available

API rate limits:

• Don't run multiple OpenClaw sessions/agents simultaneously unless they use separate API keys
• Avoid processing very large files or contexts in one shot — break them up
• Monitor your Anthropic Console usage dashboard regularly

Frequently Asked Questions

How do I know which rate limit I'm hitting?

Look at the error message. If it mentions the registry or skill installation, it's the ClawHub registry limit. If it mentions Anthropic, api.anthropic.com, rate_limit_error, or appears during a chat/agent task, it's the Claude API limit.

How long does the rate limit last?

For the ClawHub registry: typically 60–120 seconds. Heavy sustained usage may extend this to 5 minutes. For the Anthropic API: usually 1–5 minutes for standard limits. Long-context quota exhaustion may need a longer wait.

Does logging in help with the registry limit?

Yes. Authenticated requests have a higher rate limit threshold than anonymous ones. Run clawhub login to get a session token.

I only ran one command — why did I get rate limited?

A single clawhub update --all or an install of a large skill bundle can trigger multiple internal API calls (metadata fetch, file download, dependency resolution). This all counts toward your limit. Also, if your CLI is outdated, it may be making more requests than necessary.

I'm on a VPS and I keep hitting limits even after waiting.

This is a known issue with VPS and cloud server environments. Your server's IP address is often shared with many other users running similar tools, and the registry may be aggressively rate-limiting the entire IP block. Try authenticating with clawhub login, update to the latest CLI, or consider running installs from a different environment.

Can I self-host the registry to avoid rate limits entirely?

Yes. The --registry flag or CLAWHUB_REGISTRY environment variable lets you point the CLI at your own registry instance with no public rate limits.

Still seeing the error after 5+ minutes?

If it's a registry error, this may indicate a temporary outage on ClawHub.ai rather than a rate limit. Browse and discover skills on ClawHub Skills Lib while waiting — no API calls required. You can also check the ClawHub GitHub issues for ongoing reports.