Spaces:
Running
Running
Commit History
-
Skill icons: show the spinner while painting, not a stale reused-id icon
Forged skill ids live in a reused 9000+ band, so getSkillIcon(id) could return a PREVIOUS skill's art for a brand-new pending one — the loading spinner never showed; the old icon did. Two layers of fix: - paintSkillIcon / skill detail / forge card now check iconPending FIRST and show the spinner (clearing any background image), only fetching the stored icon once it's no longer pending. - forgeSkillForHero deletes any existing icon for the (reused) id up front, so the store starts clean — new delSkillIcon() in personaStore. Also: Replay tour button on the hero-selection screen (only when no hero is picked yet) → tut.reset() restarts the guided tour. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>e46cfbf -
Skill forge: usable immediately, character-in-skill art painted in background
Forging a skill no longer blocks on image generation. The coding model defines the skill, it's persisted + learned at once, and a single scene of THE CHARACTER using the skill is painted in the background (one image now, used as both icon and detail art — no separate icon + action art). - skillForge.js: persist with iconPending=true and return immediately; generateSkillImage() runs fire-and-forget, passing the hero's portrait to Klein as an identity ref so the actual character appears in the scene. Status names the coding model (like image/voice backends do). - app.py: /portrait accepts ref_image; both the sidecar path (_remote_klein_portrait) and in-process ZeroGPU path (_klein_portrait, via new _decode_ref_image) feed it as FLUX.2 identity conditioning. - imagen.js / imagenServer.js: thread refImage through to local + Klein engines. - tiny.js / skillForgePanel.js: gold spinner loading indicator while an icon is still painting; swaps to the art when iconPending clears (on…cedb3f2
-
Learn & manage skills from the in-Game character sheet
Extract the forge pipeline into shared skillForge.js (forgeSkillForHero: model → validate → Klein icon → persist), used by both the Skill Forge tab and the character sheet. On the Game character sheet add: a 3-slot Skill Bar (tap a slot to assign/clear a learned skill), an inline ★ equip toggle on each forged-skill card, and a '⚒ Forge a new skill' button that forges for the active hero in a bottom-sheet. Verified end-to-end (real forge produced a valid skill + Klein icon; slot picker + equip persist). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>727efc9 -
Phase 2: Skill Forge writes real engine skills + character-page detail screen
skillSchema.js validates/clamps model output into an engine-resolvable CB skill (safe op/condition whitelist, 9000+ ids, effect summaries, equip resolution). The forge panel now emits that IR, validates with one retry + fallback, paints a Klein icon, and persists the skill onto the hero (customSkills/learnedSkills). The Game character sheet lists forged skills as cards; tapping opens a detail bottom-sheet with a lazily-painted Klein action illustration of the hero performing the skill, flavor, effects, and equip/unequip (max 3). Equipped skills flow into live combat. Rebuilt engine/combo/sandbox bundles for the object-or-id skill bar. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>9fe1971
-
Skill Forge: optional "show thinking" for coding models
Add a "show model thinking" toggle (off by default) that reveals the reasoning models' trace in a collapsible debug panel, mirroring the persona debug div and reusing stripThink/stripThinkFinal. A `think` flag flows from the checkbox -> streamCoding -> /text/generate/stream -> the model. Reasoning comes back inline as <think>…</think>; the clean answer shows in the output, the raw trace in the panel. - Nemotron (_nim_text_stream): when think=true, drop reasoning_budget=0 and surface reasoning_content wrapped as <think>…</think>. - BLS sidecar: new 5th `think` arg streams the reasoning wrapped in <think> instead of discarding it; threaded through _space_text_stream's *extra (defaults keep existing 4-arg callers on the clean path). - Mellum2 has no reasoning, so the flag is a no-op there. Verified end-to-end through /text/generate/stream for both models (think on/off) plus the backward-compat 4-arg sidecar call. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.co…6e155d8
-
Add Coding Model setting + Skill Forge sandbox
New "Coding Model" settings section (below Persona Prompt) selecting between two backends: • Nemotron 3 Nano 30B-A3B via NVIDIA NIM (integrate.api.nvidia.com, reasoning off) — too large to self-host on ZeroGPU. • Mellum2 12B-A2.5B via a ZeroGPU sidecar (polats/tiny-army-mellum-zerogpu), same /generate(+_stream) contract as the Aya/MiniCPM5 sidecars. New "Skill Forge" tab (Sandbox group, after Sprite Animations): pick a recruited hero, describe a skill, and the chosen coding model authors a tailored skill definition. Reuses personaStore (roster) + streamCoding. Backend: app.py routes mellum2-zerogpu and nemotron-3-nano-30b-nim model ids in /text/generate/stream; _nim_text_stream() streams NIM's OpenAI-compatible SSE. Frontend: codingModel.js store (separate from the persona/diary model), codingModelBar.js picker, shared sseText.js SSE reader (extracted from engineServer.js). Needs Space config: TINY_MELLUM_SPACE + NVIDIA_NIM_API_KEY. Co-Authored-By: Claude Opus 4.8 (1M c…1f1908e