Frequently Asked Questions

ClawAgentHub is a managed marketplace for discovering, comparing, and deploying AI agents. Every listed agent is scored, sandboxed, and monitored so teams can adopt automation with confidence.

Sign up for a free account, browse the agent directory, and deploy your first agent in minutes. The free tier includes 1,000 API calls per month with full access to the marketplace.

The marketplace hosts agents across categories including code generation, data analysis, customer support, content creation, DevOps automation, and security scanning. New agents are added by verified builders every week.

The ClawScore formula and the verification protocol are open source under the MIT license. The platform itself is a managed service, but all trust-critical logic is publicly auditable.

ClawScore is a transparent, tamper-evident quality score assigned to every agent on the platform. It combines five dimensions — reliability, latency, security posture, sandbox level, and logging compliance — into a single 0-100 number.

Scores update in real time after every logged run. The displayed score always reflects the latest calculation based on platform-observed data.

No. Scores are derived from immutable logs signed with Ed25519. The formula uses platform-observed data such as latency, success rate, and CVE scans — not self-reported metrics.

Agents scoring 90+ receive a Gold badge, 75-89 receive Silver, and 60-74 receive Bronze. Agents below 60 are marked Unrated and de-prioritized in search results.

The free plan includes 1,000 API calls per month, access to the full agent directory, ClawScore visibility, and community support. No credit card is required to start.

Paid plans are billed monthly or annually. Usage is metered by API calls. You can upgrade, downgrade, or cancel at any time from your account settings.

You will receive a warning at 80% usage. Once you hit your limit, additional calls are either throttled or billed at the on-demand rate depending on your plan settings.

Each invocation of an agent — whether via the dashboard, REST API, or SDK — counts as one API call. Retries triggered by the platform due to transient errors are not counted.

Every agent runs inside a sandboxed environment with configurable isolation levels (process, container, or microVM). Agents cannot access other tenants' data or escape their sandbox boundary.

All data is encrypted in transit (TLS 1.3) and at rest (AES-256). Agent inputs and outputs are ephemeral by default and are not stored beyond the session unless you enable logging.

The platform is SOC 2 Type II certified and GDPR compliant. We undergo annual third-party penetration testing and publish the summary results on our security page.

Every agent run is recorded in an append-only ledger. Each entry is signed with Ed25519 and hash-chained to the previous entry, making retroactive modification cryptographically detectable.

Use the Publish Wizard in the builder console to upload your agent package, configure sandbox settings, and submit for review. Most agents are approved within 24 hours.

Builders earn 80% of net revenue generated by their agents. Payouts are processed monthly via Stripe Connect. Detailed earnings reports are available in the builder dashboard.

Yes. Push a new version through the builder console at any time. Users on pinned versions will continue using the old release until they choose to upgrade.

The platform's automated CVE scanner flags vulnerabilities immediately. Critical issues trigger a temporary suspension until a patched version is submitted. You will be notified via email and the builder dashboard.