Compare · Known Agents
Skald vs Known Agents: naming AI agents, and proving which are real
Skald vs Known Agents compares two tools that track AI agents on your site. Known Agents takes events from your server, CDN or WordPress and keeps your robots.txt current. Skald reads your log, checks each crawler's address, and joins the visits AI sends to your leads.
Skald vs Known Agents compares two ways to see crawlers, scrapers and AI agents. Known Agents is a hosted service fed by events from your server, CDN or WordPress. Skald reads your server's access log, sorts every request into five groups that add up to the log, checks crawlers against their owners' published addresses, and counts every figure twice.
30-day free trial · no card · we set it up within one working day · free to self-host
- 09:14:01 203.0.113.24 "GET /pricing/" "GPTBot/1.2"verified · OpenAI's list
- 09:14:03 203.0.113.9 "GET /about/" "ClaudeBot/1.0"verified · Anthropic's list
- 09:14:04 198.51.100.90 "GET /docs/" "CCBot/2.0"not verifiable
- 09:14:07 192.0.2.200 "GET /wp-login.php" "GPTBot/1.2"claims to be GPTBot
- 09:14:09 198.51.100.23 "GET /docs/" ref "chatgpt.com"a person, from ChatGPT
Sample lines. A name in a request is a claim. Skald checks it against what the crawler's owner publishes.
[01] question by question
Skald vs Known Agents, question by question
Pick what you need to know. Where Known Agents' own pages did not say, the cell says so; we never fill a gap with a guess. Every comparison is on Skald alternatives and comparisons.
What do you need to know? pick one
AI agents and crawlers: Known Agents' focus. It answers 4 of these 8 in full and 1 in part; for 3 its own pages did not say. Skald answers all 8.
| The question | Known Agentsevents from your server, CDN or WordPress | Skaldyour server log |
|---|---|---|
| Which AI training crawlers read your pages? (GPTBot, ClaudeBot) | Yes | Yes, page by page |
| Did an assistant fetch a page live, for someone? (ChatGPT-User) | Yes | Yes, kept apart from training crawls |
| Which search engine crawlers came? | Yes, "every crawler, scraper and AI agent" | Yes, in their own group |
| Block AI crawlers | Yes, an auto-updating robots.txt; WordPress rule enforcement on paid plans | Yes, by your web server's rules; it does not charge |
| robots.txt rules beside what each crawler fetched | It writes robots.txt for you | Yes, and fetches of forbidden paths per crawler |
| Is that crawler who it claims to be? | Not confirmed in its own pages | Yes, its owner's published addresses or DNS |
| Fake crawlers filed apart ("claims to be GPTBot") | Not confirmed in its own pages | Yes, under bots |
| Why did the AI come: training, a live question, or a click? | Not confirmed in its own pages | Yes, three reasons, never added together |
| People who clicked through from an AI assistant | Yes | Yes, by assistant; a floor, as apps strip referrers |
| Visitors with an ad-blocker | Yes, server-side | Yes, every request is in the log |
| Pages a CDN served from its cache | Yes, CDN connectors | Yes, from Cloudflare, CloudFront or Fastly logs |
| Visitors and visits | Visits from AI chats | Yes, every source, as an upper bound on people |
| Works where you cannot read the server's log | Needs a supported CDN, platform or server code | People only, with the script tag; no AI or bots |
| Search queries, impressions, position | No | Synced from Search Console and Bing every 6 hours |
| Events inside the page (clicks, sign-ups) | No | Yes, with the optional script tag |
| Funnels | No | Yes, counted every day by source and device |
| Retention cohorts | No | Yes, by week, split by first source |
| Leads by source and AI assistant | No | Yes, first and last touch, from signed webhooks |
| Totals must add up, or the day is withheld | No | Yes |
| A second, separate recount of every figure | No | Yes, every day |
| Data-quality panel (unreadable lines, gaps, duplicates) | No | Yes |
| Runs on your own server | No, a hosted service | Yes, free to self-host |
| Raw requests and addresses | CSV on paid plans | Yes, if you choose to keep them |
| Named users and roles | Team members on paid plans | Yes, owner, admin and viewer |
| API | On its Enterprise plan | Yes, a read-only JSON API and an MCP server |
| Several sites in one place | Yes, projects | Yes, with a site switcher |
What each tool is built to do, as we read their own docs in October 2026. "Not confirmed" means its own pages did not say either way. If we have yours wrong, tell us.
[02] to be fair
Where Known Agents does more than Skald
Known Agents is built only for crawlers, scrapers and AI agents. Here is where it may suit you better.
robots.txt kept for you
It writes an auto-updating robots.txt. Skald suggests robots.txt lines to copy, and never writes the file itself.
Rules on WordPress
On paid plans it enforces rules on WordPress. Skald's rules act through nginx on the server Skald runs beside.
Nothing to run
It is a hosted service, fed by your server, CDN or WordPress. Skald runs on a server: yours, or ours if we host it.
[03] proving the names
Is that AI agent real? Three answers, each with its proof
Anyone can write "GPTBot" in a request. Skald checks each crawler's address against the lists or DNS its owner publishes, and files every one in one of three states.
one line · sample
203.0.113.9 "GET /about/" "ClaudeBot/1.0"
AI · training · verified
The address is on Anthropic's published list. Counted as AI, under the training crawl.
The lists come from OpenAI, Anthropic, Perplexity, Google, Bing, Apple, DuckDuckGo and Meta, fetched once a day. A list older than 7 days is no longer trusted to call anything a fake.
[04] beyond the agents
Block training crawlers, stay in AI answers, and see what it brings
Skald's crawler rules work per AI company: allow it, block only its training crawlers, or block it. robots.txt always stays reachable, and every refusal is counted.
Rule
Keep GPTBot out and let ChatGPT-User fetch for a person who asked. Your web server applies it, by the name each crawler sends.
Watch
Per crawler: requests, bytes, answer time, errors and what it fetched that robots.txt forbids.
Follow
A person who clicks through from an AI answer is followed to the lead they become, with what the assistant read first.
[05] choose by the question
Known Agents, Skald, or both
- "Keep our robots.txt current without thinking about it." Known Agents.
- "Is that GPTBot really OpenAI's?" Skald.
- "Did a page an assistant read bring a lead?" Skald.
- "Make AI crawlers pay." Neither: Cloudflare charges, Skald blocks.
What it does not do
- Skald blocks crawlers; it does not charge them.
- Its rules work by the name a crawler sends, through nginx on the server Skald runs beside. A scraper behind a browser name is not stopped.
- Some crawlers are robots.txt only. A robots.txt line is a request, not a refusal.
- Skald never writes your robots.txt. It suggests the lines; you add them.
keep reading
More on Skald vs Known Agents
questions
Skald vs Known Agents FAQ
Straight answers, including what Skald does not do.
What does Known Agents do?
As we read its own pages in October 2026, Known Agents tracks every crawler, scraper and AI agent from events your server, CDN or WordPress sends it, shows visits from AI chats, writes an auto-updating robots.txt, and enforces rules on WordPress on paid plans. It is a hosted service.
How does Skald tell a real GPTBot from a fake?
It checks the address against the list or DNS the crawler's owner publishes: OpenAI, Anthropic, Perplexity, Google, Bing, Apple, DuckDuckGo and Meta. A request that uses GPTBot's name from an address OpenAI does not use is filed under bots as "claims to be GPTBot". How AI crawler checks work.
Can Skald block AI crawlers?
Yes, through your web server's rules: per AI company, allow, block only its training crawlers, or block it, with robots.txt always reachable. Skald blocks; it does not charge. How blocking AI crawlers works.
Does Known Agents count people too?
It shows visits from AI chats. Skald counts every person who visits, from any source, as an upper bound on people, with AI assistants as one source among many.
Can I run Skald on my own server?
Yes. Skald is source-available under the Elastic License 2.0 and free to self-host. Known Agents is a hosted service.
Can I use both?
Yes. Known Agents takes events from your server or CDN; Skald reads your log. Neither needs the other, and both can see the same traffic.
$ tail -f access.log | skald
See what AI does with your site. Free for 30 days.
No card. We set up hosted Skald for you within one working day, or you self-host it free. Every line of your log, sorted and added up.