replen 1.0.7 → 1.0.8

package/extras/skills/replen/SKILL.md

@@ -214,14 +214,18 @@ that actually fit."* If the user agrees, run this checklist:
      a Python CCXT market-making engine:
      `["crypto","trading","market-making","ccxt","quant","backtesting"]`).
    - **Set technical capabilities** with `replen_set_capabilities` — short,
-     GitHub-searchable tech terms for what the project DOES at the tech level
-     (e.g. `["crypto exchange","market data","backtesting","technical
-     analysis"]`; for a defense CV pipeline `["computer vision","object
-     detection","satellite imagery","geospatial mapping"]`). Derive these from
-     the actual imports/deps, not guesses. This is the highest-leverage step:
+     GitHub-searchable tech terms for what the project DOES at the tech level.
+     Aim for **8-15** and be **SPECIFIC** — specific capabilities match far
+     better than broad ones. Break a broad capability into the concrete
+     techniques the code actually uses: not just `"web scraping"` but
+     `["web scraping","headless browser","cloudflare bypass","proxy rotation",
+     "session handling","rate limiting"]`; not just `"trading"` but
+     `["crypto exchange","market data","backtesting","technical analysis",
+     "order management","websockets"]`. Derive them from the actual
+     imports/deps and code, not guesses. This is the highest-leverage step:
      the server builds the project's facet vectors from these IMMEDIATELY (no
-     waiting for a scheduled run), and they drive both faceted matching and the
-     shared capability catalogue.
+     waiting for a scheduled run), and they drive faceted matching against the
+     shared library catalogue.
 
    **Do NOT tell the user to set tags/capabilities on the web** — that's the
    sticky step this replaces; set them with the tools. (They can fine-tune later

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "replen",
-  "version": "1.0.7",
+  "version": "1.0.8",
   "description": "Make your AI coding tools smarter. One command, no API keys. Replen scouts the OSS firehose against your projects and surfaces drop-in libraries, ideas to port, and dead deps to swap — the match decision happens inside your AI tool's session on your subscription tokens. 1-3 actionable matches a month, by design.",
   "type": "module",
   "bin": {