@relvehq/agent-ready 0.3.0 → 0.4.0

package/README.md

@@ -104,7 +104,7 @@ frameworks" below. For WordPress, the plugin ships inside this package
 | `enabled` | `AGENT_READY_ENABLED` is `1` or `true` |
 | `exclude` | `/api`, `/_next`, auth and admin paths, plus common private pages (account, checkout, cart, dashboard, settings, billing…) and yours; `excludeCommon: false` drops the common list |
 | `llms` | `/llms.txt` built from your `sitemap.xml`; `llms: false` for none |
-| `onHit` | sends each visit to a dashboard when `AGENT_READY_KEY` and `AGENT_READY_ENDPOINT` are set |
+| `onHit` | sends each visit's stats to Relve when `AGENT_READY_KEY` is set (opt-in; `AGENT_READY_ENDPOINT` for another dashboard), see [Share visit stats](#share-visit-stats-with-relve-opt-in) |
 
 Anything you pass wins:
 
@@ -186,9 +186,13 @@ Markdown. The command saves `<out>/<path>.md` plus `<out>/<path>.json`
 (default out: `.agent-ready/md`).
 
 - **Cost:** your own OpenAI usage, one model call per changed page (roughly
-  a few thousand tokens each with the default `gpt-4o-mini`). Relve charges nothing.
+  a few thousand tokens each with the default `gpt-4o`, about 1 to 2 cents a
+  page). Relve charges nothing.
 - **Your key:** used for that call only. Relve does not store or log it, and
   it is never written to the output files.
+- **What Relve logs:** for each call, the page URL, the model, the outcome
+  and the size of the result (bytes, words, lines, tokens, headings, links,
+  tables). Not the HTML, not the Markdown, not your key.
 - **Re-run when content changes.** Each saved `.json` holds the page's
   fingerprint; it is sent along, and an unchanged page comes back as
   `unchanged` with no model call (free). `--force` rewrites everything.
@@ -198,7 +202,7 @@ Markdown. The command saves `<out>/<path>.md` plus `<out>/<path>.json`
 | Option | |
 |---|---|
 | `--sitemap` / `--paths /a,/b` | which pages (sitemap at `<site>/sitemap.xml`) |
-| `--model` | `gpt-4o-mini` (default, fastest), `gpt-4o`, `gpt-4.1`, `gpt-4.1-mini` |
+| `--model` | `gpt-4o` (default, most accurate), `gpt-4.1`, `gpt-4.1-mini`, `gpt-4o-mini` (cheapest; can mix up whose price is whose on busy pages) |
 | `--site-name` | your brand name, used in the documents |
 | `--out`, `--limit`, `--concurrency` (2), `--force`, `--dry-run` | `--dry-run` lists the pages: no API call, nothing written |
 | `AGENT_READY_REFINE_ENDPOINT` | another API URL (testing) |
@@ -287,6 +291,23 @@ writes a log that `npx agent-ready stats --log <file>` summarises. On
 serverless hosts (Vercel, Amplify) memory and local files don't last between
 requests: implement `HitStore` on your database.
 
+### Share visit stats with Relve (opt-in)
+
+```
+AGENT_READY_KEY=ar_live_yoursite
+```
+
+With that one variable set, `createAgentReady` reports each Markdown request
+an agent makes to `https://relvehq.com/api/agent-ready/v1/hits`: your key, the
+package version, the time, host, path, how the Markdown was asked for, where
+it came from, status, tokens, bytes, the agent's user agent and the time
+taken. Never the Markdown and never the visitor's IP. The report is
+fire-and-forget with a 2 second timeout and never changes the agent's
+response. The key is a label you choose (`ar_` plus 3 to 64 letters, digits,
+`_` or `-`), not a secret. Without it nothing is sent.
+`AGENT_READY_ENDPOINT` sends the same reports to another dashboard instead;
+in code, `reporter({ key, endpoint? })` is the `onHit`.
+
 ## Other frameworks
 
 Express, Connect, Polka, Fastify (via `@fastify/middie`) or a plain `http`