boomcode 0.5.6__tar.gz → 0.5.7__tar.gz

{boomcode-0.5.6 → boomcode-0.5.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: boomcode
-Version: 0.5.6
+Version: 0.5.7
 Summary: BOOMCODE — AI coding CLI for Termux. Your app, BLAM, done!
 Author: zyvo9
 License: MIT

{boomcode-0.5.6 → boomcode-0.5.7}/boomcode.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: boomcode
-Version: 0.5.6
+Version: 0.5.7
 Summary: BOOMCODE — AI coding CLI for Termux. Your app, BLAM, done!
 Author: zyvo9
 License: MIT

{boomcode-0.5.6 → boomcode-0.5.7}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "boomcode"
-version = "0.5.6"
+version = "0.5.7"
 description = "BOOMCODE — AI coding CLI for Termux. Your app, BLAM, done!"
 readme = "README.md"
 requires-python = ">=3.8"

{boomcode-0.5.6 → boomcode-0.5.7}/python/boomcode/__init__.py

@@ -6,4 +6,4 @@ hands over to the real installer (install.sh) which does everything
 else (core engine, deps, single-line live progress).
 """
 
-__version__ = "0.5.6"
+__version__ = "0.5.7"

{boomcode-0.5.6 → boomcode-0.5.7}/python/boomcode/layer/config/agent/build.md

@@ -111,6 +111,13 @@ You are text-only — you cannot see images directly. **Delegate EVERY image-vie
 
 **Never** invent image content you could not actually see. If vision fails, tell the user honestly that you could not see it.
 
+## Skills — follow the whole workflow
+When a skill defines a step-by-step flow, you MUST follow ALL of its steps in
+order — never shortcut it, never "just build something quickly". The skill IS
+the workflow. Extra material from the user (a photo, a link, a file) feeds
+INTO the skill's steps (as a reference for prompts, etc.) — it never replaces
+the steps and never lets you skip ahead.
+
 ## Installs — always ask first
 If anything is missing or an install becomes needed (pkg/pip/npm/apt, downloaded
 tools — anything), follow the **ask-install** skill: tell the user what is missing

{boomcode-0.5.6 → boomcode-0.5.7}/python/boomcode/layer/skills/lets-scroll/SKILL.md

@@ -27,6 +27,33 @@ world — scene after scene, no cuts — ending at the brand.
 > GUIDE.md, file names) uses the plain `image-…` / `video-…` vocabulary below —
 > never internal words like "still", "dive", "connector", "leg".
 
+## MANDATORY FLOW — no shortcuts, ever
+
+**If this skill is loaded, you follow ALL of it, in order.** The flow is:
+interview → design → write the kit (`lets-scroll/` with prompts + GUIDE.md) →
+WAIT for the user's renders → verify → assemble → check together.
+
+**You NEVER (these are the classic mistakes — do not make them):**
+- build the site directly (writing index.html first, "to save time")
+- skip the interview or the kit — the kit IS the deliverable of your design
+- render or fake assets yourself — you cannot generate images/videos; the
+  USER renders them in their own tools, always
+- copy a user-supplied photo into the page as if it were a scene — a product
+  photo is a REFERENCE, not a scene
+- jump ahead to Step 4/5 before every kit file has its render
+
+If you ever feel like shortcutting: stop, re-read this block, and write the kit.
+
+**If the user gives you a picture (e.g. their product photo):**
+1. Look at it first with `boomcode-vision` — describe it in detail (shape,
+   colors, materials, text/logo on it).
+2. That description goes INTO your prompts: every scene that features the
+   product must describe the SAME product so the generated world matches it.
+3. The photo itself stays as a reference — it does not replace any kit file
+   and does not skip any step.
+4. Then continue the normal flow (a short interview is still needed: camera
+   style, scene count, phone version).
+
 ## The idea in one minute
 
 1. You interview the user about their business/idea (simple questions).