@kenkaiiii/ggcoder 5.71.0 → 5.73.0

package/assets/motion/skills/app-walkthrough/SKILL.md

@@ -11,6 +11,10 @@ specific to walkthroughs.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 - "Which task should the video walk through?" Offer the two or three tasks the
   prompt or screenshots suggest, plus "Let GG Motion choose".
 - "Where will people mostly watch this?" (from the `motion` skill).

package/assets/motion/skills/before-after/SKILL.md

@@ -10,6 +10,10 @@ skill describes; this page adds what is specific to comparisons.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 - "What's the annoying part of life without it?" Offer two or three guesses
   from the prompt, plus "Let GG Motion choose".
 - "Where will people mostly watch this?" (from the `motion` skill).

package/assets/motion/skills/dev-tool-video/SKILL.md

@@ -10,6 +10,10 @@ describes; this page adds what is specific to developer tools.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 - "What's the one command or line of code that shows it best?" Offer the
   README's quick-start as recommended, plus "Let GG Motion choose".
 - "Where will people mostly watch this?" (from the `motion` skill).

package/assets/motion/skills/launch-video/SKILL.md

@@ -10,6 +10,10 @@ skill describes; this page adds what is specific to launches.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 Pick what the prompt leaves open, at most four questions with the idea pitch:
 
 - "Where will people mostly watch this?" (from the `motion` skill).

package/assets/motion/skills/match-reference/SKILL.md

@@ -11,6 +11,10 @@ skill describes; this page adds the study step.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 - "What should the new video be about?" when the prompt doesn't say.
 - "What do you like most about the example?" The pace · The look · How one
   scene flows into the next · All of it (recommended).

package/assets/motion/skills/motion/SKILL.md

@@ -20,13 +20,15 @@ be corrected.
 **When to ask.** For a new video from an open prompt, ask once, before planning,
 about what only the user knows and would be costly to change later: where it
 will be watched (it sets shape, length and pace), whether it has music (the
-timing is built on it) and, when the subject suggests they have some, material to
-include. Ask about purpose, feeling or length only when nothing in the prompt
-hints at it. Own everything else: look, colour, type, pacing, transitions.
+timing is built on it), and what the visuals should be made from: their own
+images, source material, original illustrations or a mix. A URL, logo or photo
+attachment alone does not choose that approach. Ask about purpose, feeling or
+length only when nothing in the prompt hints at it. Own the execution: colour,
+type, pacing and transitions, within the user's chosen approach.
 
 - Skip anything the prompt, attachments, a brand kit or workspace preferences
-  already answer. A detailed brief, a shot list or an edit gets no
-  questions.
+  already answer. An edit gets no intake questions. A detailed brief or shot
+  list skips questions only for choices it actually settles.
 - Use one `ask_user` card with at most four questions, each with a recommended
   answer drawn from the prompt. Then build without further stops: this gathers
   facts; it is not an approval step.
@@ -47,8 +49,12 @@ LUFS, safe zone, lower third, CTA, fps, kinetic type, loop seam. If one is
 unavoidable, explain it in the same line. If the user writes in those terms,
 answer in them.
 
-**Starting wording.** Adapt it to the prompt and pick at most three, plus the
-idea question below.
+**Starting wording.** All job skills use this bank. Keep one card, at most four
+questions total. When visual approach is unsettled, include it before an idea
+pitch or an optional purpose/feeling question; do not add a second card. Skip
+it for edits or when the brief or an explicitly requested reference style
+already settles it. Tailor the options to the source: website product photos,
+app screens, supplied footage or diagrams, not the same choices for every job.
 
 1. "Where will people mostly watch this?"
    - Scrolling on a phone: Tall video. Grabs attention right away and works with
@@ -69,9 +75,25 @@ idea question below.
      muted.
    - I'll add music myself: Silent for now, with a steady rhythm so your music is
      easy to add.
-3. "Do you have anything it should include?" (pick any): My logo or colours ·
+3. "What should we build the visuals from?"
+   - Use my images: Build around your own photos or footage; if not attached,
+     wait for you to supply them before building.
+   - Use website photos: Use individual product or lifestyle photos from the
+     supplied site, not screenshots of whole pages. Offer only for a real site.
+   - Animate illustrations: Create drawn scenes, objects and diagrams that
+     explain your subject, using its real brand and facts.
+   - Mix photos and animation: Combine real product images or footage with
+     illustrated scenes and transitions.
+
+   Mark the best fit recommended, but never silently choose an approach. Offer
+   only relevant options; rename the source option for app screens or other
+   supplied material. Missing assets are a blocker, not permission to invent
+   them. Illustrations use HTML/SVG/CSS; image generation still needs an explicit
+   user request. Their choice guides the ideas and the build, not just sourcing.
+
+4. "Do you have anything it should include?" (pick any): My logo or colours ·
    My photos or videos · Exact words or numbers · Nothing, start fresh
-4. "What should people get from it?"
+5. "What should people get from it?"
    - Understand something: Explains an idea, a process or some numbers, step by
      step.
    - Want to buy or try it: Shows what it does for them and ends with one clear
@@ -79,14 +101,14 @@ idea question below.
    - Hear some news: A launch, an event or a milestone.
    - Feel something: A mood piece, celebration or tribute, led by feeling rather
      than facts.
-5. "How should it feel?"
+6. "How should it feel?"
    - Calm and clear: Gentle movement and time to read. Nothing flashy.
    - Bold and energetic: Quick cuts, big words, a strong beat.
    - Playful: Bright and bouncy, with a bit of fun.
    - Elegant: Slow, precise and spacious, like a luxury ad.
    - Warm and personal: Soft and unhurried; lets photos and moments breathe.
    - Serious and respectful: Restrained, for sensitive or factual subjects.
-6. "How long should it be?"
+7. "How long should it be?"
    - About 10 seconds: One idea, quick to watch.
    - About 30 seconds: Room for a short story or a few points.
    - About a minute: Room to explain something step by step.
@@ -99,12 +121,13 @@ opens into the answer"), with your favourite marked recommended, plus
 "Let GG Motion choose". Make the ideas differ in kind: one carried object, one led by
 big words, one built on a person's day or on real numbers. Skip it when the
 prompt already describes the concept, for edits, and when the card would exceed
-four questions; then choose yourself.
+four questions; then choose the concept yourself within the selected visual
+approach. Never drop the unsettled visual-approach question to fit an idea pitch.
 
-Illustratively: "30-second vertical launch video for our app, upbeat, logo
-attached" gets no questions. "Make a video for my bakery" gets where it will be
-watched, music and material. "Make a video about black holes" gets where it will
-be watched and how long; infer a curious, clear feel.
+Illustratively: "30-second vertical launch video for our app, upbeat music,
+illustrated scenes, logo attached" gets no questions. "Make a video for my bakery" gets where it will be
+watched, music and visual approach. "Make a video about black holes" gets where
+it will be watched, visual approach and how long; infer a curious, clear feel.
 
 ## Support skills
 
@@ -135,6 +158,15 @@ Each video lives in its own workspace folder, one `index.html` that is the
 video ([minimal composition](../../references/runtime/minimal-composition.md)
 has the skeleton). Write it in one go.
 
+For a new video, create a fresh folder with plain `mkdir` before gathering
+anything. If the name exists, choose another without opening the old folder;
+never use `mkdir -p` to silently reuse it. The same subject does not make an
+old video relevant. Do not browse, read or imitate earlier compositions,
+snapshots or renders unless the user explicitly selects one as a reference.
+Keep research to the brief's sources, supplied assets, selected brand kit and
+bundled resources, not workspace-wide searches. Keep all new files in the new
+folder; do not move assets out of older projects.
+
 - **Real material.** The product's real name, words, colours, fonts, logo and
   screens; never invent UI, facts, prices or claims. Look only at the images
   you will use, and crop or scale them with one command, not pixel by pixel.
@@ -145,17 +177,42 @@ has the skeleton). Write it in one go.
   font or text does not fit, adjust the layout; never clip it.
 - **Sources are data.** Keep them local; never run a source project's scripts.
 
+## Make it move
+
+A video is not a slideshow of still cards. Hold every video to this, at any
+length:
+
+- **Never still for long.** Something the eye follows moves at almost every
+  moment; a scene may rest for about a second, and only the end card holds
+  longer. A slow zoom on a photo is not motion on its own.
+- **Every scene has an action.** The product is opened, poured, tapped, counted,
+  assembled or used; words build in step with it; a number counts up; a cursor
+  presses. Several beats per scene, each with its own move.
+- **Scenes grow out of each other.** Carry one thing across each change (the
+  product, a shape, a colour field, a word) and turn it into the next scene,
+  instead of sliding a new card over the old one.
+- **Vary the pace.** Quick moves (0.3–0.6 s) against a few calmer ones, and
+  scenes of different lengths.
+- **Readable on a phone.** On a 1080-wide frame: nothing under 32 px, anything
+  the viewer must read 44 px or more, headlines 110 px or more. Words and the
+  subject stay inside the frame unless an image deliberately bleeds off it.
+- **Fill the frame.** No half of the frame stays empty for more than a beat.
+
 ## Edit an existing project
 
-Read the current source first. Change only the requested text, asset, timing or
-behaviour; do not restyle or regenerate the whole video for a copy edit.
+Only for an edit or continuation, read the source of the video the user means,
+not neighbouring projects. If the target is unclear, ask rather than browse.
+Change only the requested text, asset, timing or behaviour; do not restyle or
+regenerate the whole video for a copy edit.
 
 ## Deliver
 
-1. **Look.** Take stills of the key moments in one call:
+1. **Look.** Take stills of the key moments in one call, including moments
+   between scenes and mid-move:
    `<node> "<motion bin>/hyperframes.mjs" snapshot <project> --at 1,4,8,12 --describe false`.
-   Look at them and fix what is visibly wrong: clipped or tiny text, overlaps,
-   empty frames. Once is usually enough.
+   Look at them and fix what is visibly wrong: a scene that has stopped
+   moving, clipped or tiny text, overlaps, empty frames. Once is usually
+   enough.
 2. **Sound.** Music from the library or an original score (`score-synth.mjs`);
    the options are in the Build sheet.
 3. **Render once**, the final, with motion blur, into a new versioned file

package/assets/motion/skills/website-video/SKILL.md

@@ -1,59 +1,76 @@
 ---
 name: website-video
-description: Job skill for a video about a website or landing page, built from the real site. Adds website questions, capture and rebuild steps, moves and pitfalls. Load once, before asking, when the user names a site or URL.
+description: Turn a website into an original animated brand film using its real product, facts and identity. Show the website itself only when requested. Load once, before asking, when the user names a site or URL.
 ---
 
 # Website video
 
-A website video makes people want to visit, using the real site.
+The website is source material, not the video. Ask which visual approach the
+user wants: their own images, website photos, illustrations or a mix. Build a
+brand film about what it offers in that approach, not a tour of its pages.
 Build it as the `motion` skill describes; this page adds what is specific to
-sites.
+sites. Demonstrate the website itself only when the user requests a walkthrough.
 
 ## Ask (on the motion skill's single card)
 
+Use the shared visual-approach question from `motion` when unsettled, ahead of
+an optional idea pitch. Adapt its choices to this job; keep the card to four
+questions total and skip choices the brief or requested reference already settles.
+
 - "What should people do after watching?" Visit, sign up, buy, book. Offer the
   site's own call to action as recommended.
 - "Where will people mostly watch this?" (from the `motion` skill).
 - "Should it have music?" (from the `motion` skill).
 - "Which idea should we go with?" Pitches below, plus "Let GG Motion choose".
 
-Without a URL, end with one plain line asking for the address.
+Without a URL, end with one plain line asking them to supply the address.
 
 ## Gather the site
 
-Capture the page once (`hf capture <url> -o sources/site`) and read its
-headline, sections, colours, fonts and images; they are the brand. Look at only
-the images you will use. Rebuild the parts that move (the hero, a menu, a
-form) as HTML. Never invent sections, prices or testimonials.
+Work inside the fresh video folder created by the `motion` skill. Capture the
+requested page once (`hf capture <url> -o sources/site`) there and read its
+headline, sections, colours, fonts and images; they are the brand. An earlier
+video about the same URL is not a source or design reference unless the user
+explicitly selects it. Look at only the images you will use.
 
-## Ideas that suit websites
+Page screenshots are research, not default scene backgrounds. Follow the chosen
+approach: individual source images for photo-led work, original HTML/SVG/CSS illustrations
+for illustrated work, or both for a mix. Keep the real logo and brand fonts.
+Do not sneak photo-led scenes into a request for illustrations or replace the
+user's photos with illustrations.
+Preserve real packaging and logos; illustration must not invent product
+features, ingredients, outcomes, prices or testimonials. Do not run image
+generation unless the user explicitly requests it.
 
-- **The site tells its own story.** The headline types, the hero builds, the
-  key section opens out of the hero, the button is pressed and becomes the end
-  card.
-- **Mechanisms, not pages.** Show three or four of the site's working parts as
-  close-up components, each poked by the cursor, rather than whole pages.
-- **One scroll, operated.** A single continuous camera move down the page with
-  real stops on the parts that matter.
+## Ideas from the subject, not the page layout
 
-## Beats that work
+Pitch actions specific to what the business offers, not ways to reveal a
+screenshot. Each idea needs a visual subject doing something beyond a text
+entrance; do not add a planning document.
 
-- A short, punchy setup (a few words or the problem), then the site.
-- Pages only full-frame and only with the cursor visible; otherwise close-ups.
-- One burst where the energy peaks (a few pages flying through, each held
-  under a second), never at the start.
-- A still hold on the site's best frame, then the URL.
+- **The product in action.** A real product cutout meets illustrated elements
+  grounded in the source: shapes pour, mix, unfold or assemble around it. One
+  object carries the viewer into the next scene.
+- **A change you can see.** Illustrate the problem becoming the solution: a
+  tangled path straightens, scattered pieces organise, a process completes.
+  Metaphors must not imply unsupported results or literal product behaviour.
+- **A small world around the brand.** Build a scene from its subject, colours
+  and shapes. Objects interact with words and transform into the final product
+  and call to action. Do not default to floating decorative blobs.
 
-## Moves
+## If a website walkthrough is requested
 
-`kit.typewrite` for the headline; `kit.cursorPath` for clicks; `kit.diveThrough`
-into a card or section; `kit.reshape` from button to end card;
-`kit.handheld` during the scroll. Pieces: `browser-window`, `zoom-into-card`,
-`request-to-result`, `cursor-click`, `camera-rig`.
+Use real UI and reproduce the specific interaction being explained as HTML.
+Show a cursor only where it operates something; use close-ups readable at the
+output size. Carry a selected component into the next scene rather than
+cycling through screenshots. A screenshot is appropriate evidence of the
+interface, not a substitute for the action.
 
 ## Pitfalls
 
-- A gallery of page thumbnails with a slow dolly: a slideshow.
-- Every page at the same length.
-- Page after page replacing each other with no handoff.
+- Website screenshots with animated text over them: not a brand film.
+- Product photos on successive cards with only zooms, fades or text entrances.
+- Following the site's section order instead of telling a story about its subject.
+- Browser frames, scrolling and cursors when the brief is not a walkthrough.
+- Decorative motion unrelated to the product, or invented visual claims.
 - Text from the site shown too small to read.

package/assets/skills/bulletproof/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: bulletproof
-description: Use when writing or reviewing code an attacker can reach — auth, sessions, tokens, crypto; untrusted input (requests, files, archives, repo content, model/tool output); secrets; per-user or multi-tenant data access (incl. Supabase/Firebase RLS); dependencies, install scripts, CI/CD, publishing, signing, updates; shelling out, deserialization, dynamic loading; LLM/agent/MCP tools; and pre-ship "can this be hacked" reviews, hardening, or suspected compromise. Fires mid-build on adding a login, query, upload, webhook, dependency, or tool. Do NOT use for throwaway scripts with no untrusted input or secrets, styling/copy/docs, or legal/privacy duty questions (compliance-guard owns those).
+description: Use when writing or reviewing code an attacker can reach — auth, sessions, tokens, crypto; untrusted input (requests, files, archives, repo content, model/tool output); secrets; per-user or multi-tenant data access (incl. Supabase/Firebase RLS); dependencies, install scripts, CI/CD, publishing, signing, updates; shelling out, deserialization, dynamic loading; LLM/agent/MCP tools; and pre-ship "can this be hacked" reviews, hardening, or suspected compromise. Fires mid-build on adding a login, upload, webhook, dependency, or tool. Do NOT use for local CLIs/libraries/scripts with no auth, network, secrets or multi-user data, styling/copy/docs, or legal/privacy questions (compliance-guard).
 license: Apache-2.0. Content is defensive engineering guidance, not a security certification or a penetration test. See references/provenance.md.
 compatibility: Works offline from the bundled references, which are a snapshot dated 3 October 2026. Version numbers, CVEs, and incident details decay fast; re-verify date-sensitive claims with web access before stating them as current. Never certifies that software is secure.
 ---

package/assets/skills/lean/SKILL.md

@@ -1,6 +1,6 @@
 ---
 name: lean
-description: Use when speed or resource use matters — while writing features that load, render lists, fetch, poll, spawn processes or cache (inline gate); on "slow / laggy / eating RAM / fans spinning", leaks, zombie or orphan processes, bundle bloat, dead code, Core Web Vitals; for a performance pass or pre-ship "will this run smoothly" check. Any stack — web, backend/API/CLI, desktop (Electron, Tauri), mobile, native, game, ML. Do NOT use for copy/docs-only changes, design direction or visual polish (that is evidence-led-ui; this skill's styling scope is payload and consistency), data-loss or migration safety (durable), or when the user explicitly deprioritizes performance.
+description: Use when speed or resource use matters — while writing features that load, render lists, fetch, poll, spawn processes or cache (inline gate); on "slow / laggy / eating RAM / fans spinning", leaks, zombie or orphan processes, bundle bloat, dead code, Core Web Vitals; for a performance pass or pre-ship "will this run smoothly" check. Any stack — web, backend/API/CLI, desktop (Electron, Tauri), mobile, native, game, ML. Do NOT use for correctness bugs where speed is not the complaint, copy/docs-only changes, design or visual polish (evidence-led-ui), data-loss or migration safety (durable), or when the user explicitly deprioritizes performance.
 license: Performance engineering guidance, not a benchmark certification. Sources and snapshot date are recorded at the foot of each reference file.
 compatibility: Snapshot dated 3 October 2026. Thresholds, tool names, and defaults decay — re-verify with web access before asserting them as current. Claims sourced to that date carry a SNAPSHOT marker.
 ---

package/dist/app-sidecar.js

@@ -62,6 +62,7 @@ import { getGitHubOpenCounts, getGitHubRepoSlug } from "./utils/github.js";
 import { startGitHubCIPoll } from "./utils/github-ci.js";
 import { extractPlanSteps } from "./utils/plan-steps.js";
 import { getNextThinkingLevel, getSupportedThinkingLevels, isThinkingLevelSupported, } from "./core/thinking-level.js";
+import { parseThinkingLevel } from "./cli/thinking-arg.js";
 import { PROMPT_COMMANDS } from "./core/prompt-commands.js";
 import { loadCustomCommands } from "./core/custom-commands.js";
 import { discoverProjects } from "./core/project-discovery.js";
@@ -492,6 +493,7 @@ async function runJsonModeIfRequested() {
             tools: { type: "string" },
             "mcp-servers": { type: "string" },
             "prompt-cache-key": { type: "string" },
+            thinking: { type: "string" },
         },
         allowPositionals: true,
         strict: true,
@@ -528,6 +530,7 @@ async function runJsonModeIfRequested() {
         allowedTools,
         allowedMcpServers,
         promptCacheKey: values["prompt-cache-key"],
+        thinkingLevel: parseThinkingLevel(values.thinking),
     }).catch(async (err) => {
         process.stderr.write((err instanceof Error ? err.message : String(err)) + "\n");
         process.exit(1);