pi-cloudflare 0.6.10 → 0.7.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "pi-cloudflare",
-  "version": "0.6.10",
+  "version": "0.7.0",
   "description": "Cloudflare Agent Plugin and native Pi extension providing official skills and cf_-prefixed MCP tools.",
   "keywords": [
     "agent-plugin",

package/plugin.json

@@ -1,7 +1,7 @@
 {
   "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
   "name": "pi-cloudflare",
-  "version": "0.6.10",
+  "version": "0.7.0",
   "description": "Cloudflare skills and cf_-prefixed MCP tools, with a native Pi adapter and an Agent Plugins 1.0 portable runtime.",
   "author": {
     "name": "0xPlayerOne",

package/skills/cloudflare-hardening/SKILL.md

@@ -103,6 +103,17 @@ Then check the ruleset phases and add what is missing:
 - **`http_request_cache_settings`** — only with a measured reason; see
   [Safety rules](#safety-rules-do-not-blanket-change).
 
+And remove what hurts performance. Hardening is not only additive:
+
+- **`enable_js` (JavaScript Detections)** — injects
+  `/cdn-cgi/challenge-platform/scripts/jsd/main.js` into every HTML response and
+  trips Lighthouse Best Practices "Avoid deprecated APIs", costing roughly 40
+  Best-Practices points on a fast site. Turn it off; it protects nothing that
+  `ai_bots_protection: "block"` does not cover without injecting anything.
+- **`fight_mode` (Bot Fight Mode)** — a genuine tradeoff rather than a free win.
+  It adds edge protection but rides on the same injected detection script. Report
+  the cost and let the user choose.
+
 ### 3. Per zone: DNSSEC
 
 Read `GET /zones/:id/dnssec`. If `disabled`, it is safe to stage: enabling is

package/skills/cloudflare-hardening/references/api-reference.md

@@ -66,14 +66,43 @@ Check that field before attempting a write; the PATCH will not tell you.
 The AI-bot and fight-mode fields live here, not under `/settings`. Use `PUT` —
 `PATCH` answers `405 10405`.
 
+Send **only the fields being changed**. Echoing the whole object back fails with
+`400 Bad Request`, because the write rejects computed and read-only fields
+(`is_robots_txt_managed`, `using_latest_model`, and friends). Every other
+setting is preserved on its own, so a partial body is correct rather than risky.
+
 ```jsonc
-// read first, then write back the merged object
+// read first to see what is on
 GET /zones/:zone_id/bot_management
 
 PUT /zones/:zone_id/bot_management
 { "ai_bots_protection": "block" }   // disabled | only_on_ad_pages | block
 ```
 
+### JavaScript Detections and Bot Fight Mode inject a script
+
+Two fields on the same object add JavaScript to **every HTML response**:
+
+| Field | Dashboard toggle | Effect |
+| --- | --- | --- |
+| `enable_js` | Security → Bots → JavaScript Detections | Injects `/cdn-cgi/challenge-platform/scripts/jsd/main.js` |
+| `fight_mode` | Security → Bots → Bot Fight Mode | Adds edge protection; also drives the injected detection script |
+
+The injected script trips Lighthouse Best Practices **"Avoid deprecated APIs"**
+and costs roughly 40 Best-Practices points on a fast site, so treat `enable_js:
+true` as a performance finding rather than a security win. Turning both off is a
+tradeoff, not a free fix — it also removes edge bot protection, so say so in the
+report.
+
+```jsonc
+PUT /zones/:zone_id/bot_management
+{ "enable_js": false, "fight_mode": false }
+```
+
+`ai_bots_protection: "block"` is the compromise: it blocks AI scrapers at the
+edge, injects no script, and costs nothing. Prefer it when the user wants bot
+protection without the performance hit.
+
 ## Free managed WAF ruleset
 
 The managed ruleset object **already exists** on every zone but is not applied —

package/skills/cloudflare-hardening/scripts/audit.mjs

@@ -207,13 +207,31 @@ async function auditZone(zone) {
   }
 
   // Bot management (AI bots / fight mode live here, not under /settings).
+  // enable_js injects a detection script into every HTML response, so it is a
+  // performance finding rather than a security gap. fight_mode is a genuine
+  // tradeoff and is reported, not flagged.
   const bot = await get(`/zones/${zone.id}/bot_management`)
   if (bot.result) {
-    if (bot.result.ai_bots_protection && bot.result.ai_bots_protection !== 'block') {
-      add(zone.name, 'info', 'ai_bots_protection', `currently ${bot.result.ai_bots_protection}`, 'consider "block" (PUT, not PATCH)')
+    if (bot.result.enable_js === true) {
+      add(
+        zone.name,
+        'high',
+        'javascript detections',
+        'injects /cdn-cgi/challenge-platform/scripts/jsd/main.js into every HTML response',
+        'turn off (PUT bot_management {enable_js:false}) — it trips Lighthouse Best Practices "deprecated API"'
+      )
     }
-    if (bot.result.fight_mode === false) {
-      add(zone.name, 'info', 'bot fight mode', 'disabled', 'consider enabling')
+    if (bot.result.fight_mode === true) {
+      add(
+        zone.name,
+        'info',
+        'bot fight mode',
+        'on — a real tradeoff, not a free win',
+        'adds edge protection but costs roughly 40 Lighthouse Best-Practices points via the injected detection script'
+      )
+    }
+    if (bot.result.ai_bots_protection && bot.result.ai_bots_protection !== 'block') {
+      add(zone.name, 'info', 'ai_bots_protection', `currently ${bot.result.ai_bots_protection}`, 'consider "block" — enforced at the edge, injects no script')
     }
   }