@agentskillshub/cli 0.2.4 → 0.2.5

package/README.md

@@ -37,7 +37,7 @@ Add `--json` to any of `search` / `audit` / `install` for machine-readable outpu
 
 ## How it works
 
-The catalog (~20K quality skills, stars ≥ 5) is a single static file (~1.7MB gzipped) served from the CDN. `ash` downloads it once, caches it at `~/.cache/agentskillshub/`, and refreshes every ~8h. **All searching is local** — fast, works offline, and puts zero load on the backend.
+The catalog (~20K quality skills, stars ≥ 5) is a single static file (~1.7MB gzipped) served from the CDN. The CLI downloads it once, caches it at `~/.cache/agentskillshub/`, and checks for a newer index with a cheap 77B probe (re-downloading the full file only when it actually changed, so a freshly-deployed index reaches you within minutes). **All searching is local** — fast, works offline, and puts zero load on the backend.
 
 ## Security grades
 

package/SKILL.md

@@ -7,7 +7,7 @@ description: Use when the user wants to find, evaluate, audit, or install an ope
 
 Discover → audit → install open-source AI agent skills and MCP servers without leaving the terminal. Backed by [AgentSkillsHub](https://agentskillshub.top): ~106K indexed skills, of which ~20K (stars ≥ 5) are in the searchable catalog, each carrying a **quality score** and a **security grade**.
 
-The catalog is a static index downloaded once and cached locally (`~/.cache/agentskillshub/`, refreshed every ~8h). Every search after the first is **instant, offline, and puts zero load on the Hub backend**.
+The catalog is a static index downloaded once and cached locally (`~/.cache/agentskillshub/`), with a cheap 77B freshness probe that re-downloads only when the index actually changes. Every search after the first is **instant, offline, and puts zero load on the Hub backend**.
 
 ## When to use
 

package/bin/ash.mjs

@@ -26,12 +26,18 @@ import { join } from "node:path";
 const BASE = process.env.AGENTSKILLSHUB_BASE || "https://agentskillshub.top";
 const META_URL = `${BASE}/search-index-meta.json`;
 const INDEX_URL = `${BASE}/search-index.json.gz`;
-const HUB_SKILL = (full) => `${BASE}/skill/${full}/`;
+// ?ref=cli lets web analytics attribute clickthroughs that originate in the CLI
+// (anonymous, no in-CLI telemetry). Query comes before any #audit fragment.
+const HUB_SKILL = (full) => `${BASE}/skill/${full}/?ref=cli`;
 
 const CACHE_DIR = join(process.env.AGENTSKILLSHUB_CACHE || join(homedir(), ".cache", "agentskillshub"));
 const CACHE_INDEX = join(CACHE_DIR, "search-index.json");
 const CACHE_META = join(CACHE_DIR, "search-index-meta.json");
-const TTL_MS = 8 * 60 * 60 * 1000; // refresh at most every 8h (matches sync cadence)
+// Within this window we serve the cached index without even probing the CDN.
+// Kept short so a freshly-deployed index reaches users within minutes — the
+// probe is a cheap 77B meta fetch; the 1.7MB index re-downloads only when
+// generated_at actually changed.
+const TTL_MS = 15 * 60 * 1000; // 15 min
 
 // security_grade → display
 const GRADE = {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@agentskillshub/cli",
-  "version": "0.2.4",
+  "version": "0.2.5",
   "description": "Search, audit, and install open-source AI agent skills & MCP servers from the terminal — security-graded, quality-scored.",
   "type": "module",
   "bin": {