Frequently Asked Questions

Kerq is a trust intelligence layer for the AI tool ecosystem. It measures and scores the real-world reliability of AI tools — APIs, plugins, MCP servers, and integrations — so your agents can make smart, trustworthy connections.

Instead of blindly connecting to any tool a user requests, your agent can query Kerq first, get a trust score, and decide whether the tool meets your reliability bar. Think of it as a credit score for AI tools.

Kerq continuously probes tools and evaluates them across 7 proprietary metrics: uptime, response time, error rate, consistency, throughput, recovery behavior, and dependency health.

Each metric is scored independently and combined into a single Kerq Score (0–100). Scores update regularly as new probe data comes in — there's no static snapshot.

The specific formulas, weights, and thresholds are proprietary. What matters to you: a higher score means a more reliable tool.

Sign up on the homepage with your email address. Once approved, you'll receive a magic login link. Visit the Developer Portal to generate and manage your API keys.

The Free tier is available immediately. High-Volume access is available for teams that need higher limits.

A standard Kerq API response includes:

• Trust score — Overall 0–100 score for the tool
• Verification status — Whether the tool has an active, verified endpoint
• Uptime percentage — Rolling availability over the measurement window
• Response time — Median latency in milliseconds
• Error rate — Percentage of failed or error-class responses

The API is designed to be machine-readable — built for agents, not dashboards.

• Free tier: 10,000 API calls/month, 20 requests/second
• High-Volume: Unlimited API calls, 100 requests/second

If you exceed your rate limit, requests return a 429 Too Many Requests response. Your monthly quota resets on your billing cycle date.

• Free: $0/month — 10k calls/month, 20 req/s. Ideal for development and low-volume production.
• High-Volume: $499/month — Unlimited calls, 100 req/s. Built for production agents running at scale.

There are no per-query charges. No surprise invoices.

Yes — this is exactly what Kerq is built for.

Before your agent connects to a new tool or MCP server, query Kerq to get the tool's trust score. If the score doesn't meet your threshold (e.g., below 70), reject the connection or fall back to an alternative. This gives your agent a principled, data-driven way to manage tool risk without hard-coding every decision.

Kerq probes tools on a rolling basis. Scores reflect recent performance — not a one-time snapshot. Tools with active endpoints are checked regularly, and scores recalculate automatically as new data comes in.

Requests beyond your limit return 429 Too Many Requests. Your quota resets on your monthly billing date. If you're consistently hitting limits on the Free tier, it's time to upgrade to High-Volume.

Log into your Developer Dashboard and click Upgrade to High-Volume. You'll be taken to a secure checkout. Your new limits activate immediately after payment.

No formal SLA at this time. Kerq is built to be reliable, but we don't offer contractual uptime guarantees yet. If that's a hard requirement, email us at info@kerq.dev to discuss your use case.

Yes. The High-Volume tier is explicitly designed for production use at scale — unlimited calls, 100 req/s, no rate surprises. The Free tier works fine for low-traffic production or staging environments.

Tools are discovered and added from public directories automatically. If your tool isn't listed yet, you can submit it directly on the Submit a Tool page — it takes about 2 minutes.

Once submitted, Kerq begins probing your tool's endpoint and generates a live trust score within 24 hours.

Find your tool in the Score Index and click Claim This Tool on the tool's detail page. You'll verify ownership by email. Once claimed, you can update your tool's details and access certification.

No. Never.

Kerq doesn't sell rankings, sponsored placements, or score boosts. The moment we did, the trust scores would become meaningless. Our business model depends on scores being real, so they will always be real.

Improve your tool's actual performance. Kerq scores the following:

• Uptime — Keep your service up. Downtime is the fastest way to tank your score.
• Response time — Faster is better. Sub-200ms responses score very well.
• Error rate — Return appropriate status codes. Consistent 5xx errors hurt significantly.
• Consistency — Low variance in response times signals a stable, well-provisioned service.
• Throughput — Can you handle load? Latency under pressure matters.
• Recovery — When you do have errors, recover quickly.
• Dependency health — Minimize disruptions from your own external dependencies.

There's no shortcut. Better infrastructure = better score.

Kerq sends lightweight, read-only HTTP requests to your tool's endpoint at a low, regular frequency. These probes are designed to have negligible impact on your service — they won't affect your performance or trip typical rate limiters.

We only record your tool's availability, response time, and status code. We never store response bodies — if your endpoint returns data, Kerq discards it immediately without reading it.

If your endpoint requires authentication, it will return a 401 or 403 — both score as reachable, since your tool is responding correctly. Authenticated probe support for claimed tools is on our roadmap.

Certification is a voluntary $499/year program that lets tools publicly display a 30-day rolling performance track record on their Kerq listing.

Certified tools get:

• A ✦ Kerq Certified badge on their listing page
• A visual 30-day performance timeline (green/yellow/gray periods)
• An embeddable badge for their website and docs

Certification tells developers: "This tool has actively chosen to be transparent about its performance." That signal matters when agents are making trust decisions.

No. Your Kerq Score is calculated identically for every tool, certified or not. Paying for certification does not boost your score by a single point.

Certification is about visibility and transparency — it gives developers confidence that you're invested in reliability. Your actual score is earned, not purchased.

The badge is granted after your tool achieves 30 consecutive days where all 7 metrics score at or above 80. Not an average — every individual metric must clear 80.

This is a meaningful bar. It's designed to reflect sustained, consistent reliability — not a lucky week.

Yes — and that's what makes it worth something.

The badge requires 30 consecutive days with all metrics at 80 or above. Your $499 subscription opens 12 months of eligibility — during that window, your tool page shows a public timeline of 30-day evaluation periods. 🟢 Green check = above 80 that period. 🟡 Yellow check = you dipped below.

If a yellow check breaks your 30-day streak, the badge is removed.

Not permanent. Get back above 80 and the countdown resets. Earn the streak again, earn the badge back.

When your annual subscription expires, your badge is removed and your performance history is hidden from your public listing. You can re-subscribe at any time — you'll start fresh and need to earn the badge again over 30 days.

You'll receive a reminder email 30 days before your renewal date.

Yes. Certified tool providers receive an embed code snippet from their listing page. The badge is a live SVG — it automatically reflects your current certification status, so it goes gray if the badge lapses and comes back purple when you earn it again.

Drop it in your README, docs, or website. It updates in real time with no maintenance on your end.

Absolutely not.

Every tool on Kerq is scored by the same objective criteria, regardless of whether they've paid anything. A free community listing and a certified tool go through the exact same scoring engine with the exact same weights.

The only thing money buys on Kerq is visibility tools (the certified badge) — and even those only work if your actual performance backs them up. You can't buy a good score. You can only earn one.

Kerq was founded by an industry veteran with extensive experience as a project manager across numerous startups, apps, and tech projects. The founder's identity is kept private to keep the focus on the product and its mission — not personalities.

Kerq discovers tools through public directories, developer communities, and curated sources. We regularly scan for new tools entering the AI/MCP ecosystem and add them automatically.

Tool providers can also submit their tools directly on the Submit a Tool page to ensure accurate information and speed up the process.

Yes. Use the Submit a Tool page — anyone can submit a tool, not just the tool's owner. If the tool has a public endpoint, Kerq will start probing it and generate a trust score.

No. Kerq does not sell user data or API usage data to third parties. Your API keys, usage patterns, and email address are used solely to operate the service.

Email us at info@kerq.dev for any questions, partnership inquiries, or feedback. We read everything.