build-small-hackathon/hackathon-advisor
16
1{2 "count": 108,3 "results": [4 {5 "slug": "dental-soap",6 "matches": [7 {8 "quest": "Backyard AI",9 "confidence": 0.96,10 "evidence": "builder's own case took three months and two specialties to untangle",11 "source": "readme"12 },13 {14 "quest": "Tiny Titan",15 "confidence": 0.97,16 "evidence": "4B open model inside the Space; Qwen3-4B-Instruct-2507",17 "source": "readme"18 },19 {20 "quest": "Off the Grid",21 "confidence": 0.9,22 "evidence": "from transformers import AutoModelForCausalLM, AutoTokenizer; no cloud client",23 "source": "app_file"24 },25 {26 "quest": "Sharing is Caring",27 "confidence": 0.85,28 "evidence": "from pdf_export import build_pdf",29 "source": "app_file"30 },31 {32 "quest": "Field Notes",33 "confidence": 0.78,34 "evidence": "Why This Exists (A Real Case) write-up of the build motivation",35 "source": "readme"36 },37 {38 "quest": "Off-Brand",39 "confidence": 0.82,40 "evidence": "header_html, footer_html, rail_html, render_handoff_html custom HTML",41 "source": "app_file"42 }43 ],44 "signal_profile": "both",45 "audit": "Kept all 6; evidence verified verbatim. Off the Grid valid: APP_FILE only local transformers, no cloud client; README's 'frontier cloud model' was the builder's PAST manual workflow, not the deployed app. Tiny Titan genuine 4B Qwen3. Tightened Off-Grid/Tiny-Titan quotes. Field Notes lowered to 0.78 (narrative, not devlog). Dropped a Best Agent add (only imports, no proven loop)."46 },47 {48 "slug": "pakistan-notice-helper",49 "matches": [50 {51 "quest": "Off the Grid",52 "confidence": 0.92,53 "evidence": "from llama_cpp import Llama; MiniCPM-V-4_6 GGUF weights",54 "source": "app_file"55 },56 {57 "quest": "Llama Champion",58 "confidence": 0.98,59 "evidence": "from llama_cpp import Llama",60 "source": "app_file"61 },62 {63 "quest": "OpenBMB",64 "confidence": 0.98,65 "evidence": "DEFAULT_MODEL_REPO = \"openbmb/MiniCPM-V-4.6-gguf\"",66 "source": "app_file"67 },68 {69 "quest": "Off-Brand",70 "confidence": 0.93,71 "evidence": "custom mobile-first frontend served by gradio.Server",72 "source": "readme"73 },74 {75 "quest": "Backyard AI",76 "confidence": 0.78,77 "evidence": "Backyard AI project; people receive convincing payment notices",78 "source": "readme"79 }80 ],81 "signal_profile": "both",82 "audit": "Kept Off the Grid, Llama Champion, OpenBMB (all literal in APP_FILE: llama_cpp import, GGUF weights, openbmb repo, no cloud API). Kept Off-Brand (custom gradio.Server, no default UI). Kept Backyard AI lowered to 0.78 (self-declared, real scam problem, but general 'people' not one named persona). DROPPED Tiny Titan: MiniCPM-V is ~8B-class; '4.6' is version not params."83 },84 {85 "slug": "Kintsugi-Garden",86 "matches": [87 {88 "quest": "Tiny Titan",89 "confidence": 0.95,90 "evidence": "lightweight instruction model (microsoft/Phi-4-mini-instruct)",91 "source": "app_file"92 },93 {94 "quest": "Off the Grid",95 "confidence": 0.78,96 "evidence": "transformers imported lazily inside load_model(); no cloud API clients",97 "source": "app_file"98 },99 {100 "quest": "Thousand Token Wood",101 "confidence": 0.82,102 "evidence": "deterministic mandala generator (PIL) that visualizes the symbols",103 "source": "readme"104 }105 ],106 "signal_profile": "both",107 "audit": "KEPT Tiny Titan (Phi-4-mini ~3.8B, literal quote). KEPT Off the Grid (local transformers, no cloud client visible; conf lowered, file truncates). KEPT Thousand Token Wood (mandala/dream art). DROPPED Off-Brand: tabs=default Gradio, no custom CSS/HTML. DROPPED Backyard AI: generic, no named person. DROPPED Field Notes: docstring is architecture desc, not a build write-up."108 },109 {110 "slug": "lolaby",111 "matches": [112 {113 "quest": "Backyard AI",114 "confidence": 0.97,115 "evidence": "kindergarten teacher runs nap time for fifteen 4-year-olds",116 "source": "readme"117 },118 {119 "quest": "Off the Grid",120 "confidence": 0.95,121 "evidence": "Everything runs locally. No cloud LLM, no per-song API cost",122 "source": "readme"123 },124 {125 "quest": "Llama Champion",126 "confidence": 0.97,127 "evidence": "from llama_cpp import Llama",128 "source": "app_file"129 },130 {131 "quest": "Tiny Titan",132 "confidence": 0.9,133 "evidence": "a small, on-device AI; lolaby-llama-3b (3B), Kokoro-82M",134 "source": "readme"135 },136 {137 "quest": "OpenBMB",138 "confidence": 0.9,139 "evidence": "MiniCPM-V can't load (openbmb/MiniCPM-V-4_6)",140 "source": "app_file"141 },142 {143 "quest": "Thousand Token Wood",144 "confidence": 0.85,145 "evidence": "watches your child's drawings and sings them a personalised lullaby",146 "source": "readme"147 }148 ],149 "signal_profile": "both",150 "audit": "KEPT Backyard AI, Off the Grid (no cloud client in app_file), Llama Champion (verbatim import), OpenBMB (MiniCPM-V in app_file), Tiny Titan (3B/82M <=4B). FIXED OpenBMB/Tiny Titan/Backyard evidence to findable text. DROPPED Well-Tuned: cited README never says fine-tune/LoRA/SFT, only repo name+tag. ADDED Thousand Token Wood: creative lullaby framing."151 },152 {153 "slug": "exam-panic-rescue",154 "matches": [155 {156 "quest": "Off-Brand",157 "confidence": 0.9,158 "evidence": "custom CSS :root variables, radial-gradient backgrounds, custom font-family",159 "source": "app_file"160 },161 {162 "quest": "Field Notes",163 "confidence": 0.8,164 "evidence": "build notes drafted in docs/codex-build-trace.md and docs/demo-script.md",165 "source": "readme"166 },167 {168 "quest": "Backyard AI",169 "confidence": 0.82,170 "evidence": "one stressed student, one exam, one time box, one final sheet",171 "source": "readme"172 }173 ],174 "signal_profile": "both",175 "audit": "Off-Brand KEPT, evidence FIXED: gr.Blocks not in truncated APP_FILE; retied to present custom CSS (:root vars, gradients, font-family). Field Notes KEPT (README links drafted codex-build-trace.md + demo-script.md), confidence trimmed (in-progress notes). Backyard AI KEPT (verbatim quote, concrete student persona). No Off the Grid: no API but no visible local model, declared_models empty. No add."176 },177 {178 "slug": "AI-agent-Evaluation-pipeline",179 "matches": [],180 "signal_profile": "none",181 "audit": "Dropped Best Agent: project EVALUATES agent traces; not itself agentic. \"Span tool-call accuracy\" is a scoring metric, \"Tool Calling demo trace\" is sample input. Off the Grid contradicted: README \"works offline\" but APP_FILE uses \"the cloud InferenceClient\". No model named so Tiny Titan/Nemotron fail; charts displayed not exportable; README is a feature list. No match survives."182 },183 {184 "slug": "global-leaders",185 "matches": [186 {187 "quest": "Thousand Token Wood",188 "confidence": 0.96,189 "evidence": "political-strategy game where a small language model runs the world",190 "source": "readme"191 },192 {193 "quest": "Nemotron",194 "confidence": 0.9,195 "evidence": "Backend: Ollama Cloud (Nemotron) if OLLAMA_API_KEY is set",196 "source": "app_file"197 }198 ],199 "signal_profile": "both",200 "audit": "KEPT Thousand Token Wood (verbatim README premise). KEPT Nemotron, FIXED evidence to app_file docstring 'Ollama Cloud (Nemotron)' (labeler cited README badge as app_file). DROPPED Off-Brand: SFX/STANCE_DOT are game content, no gr.HTML/Blocks css/theme visible. DROPPED Field Notes: README describes mechanics, not a build write-up. No Off the Grid (uses Ollama Cloud) or Tiny Titan (32B)."201 },202 {203 "slug": "pocket-weather-theater",204 "matches": [205 {206 "quest": "Well-Tuned",207 "confidence": 0.96,208 "evidence": "TUNED_ADAPTER_ID ...smollm2-135m-lora; from peft import PeftModel",209 "source": "app_file"210 },211 {212 "quest": "Tiny Titan",213 "confidence": 0.97,214 "evidence": "MODEL_ID HuggingFaceTB/SmolLM2-135M-Instruct (135M)",215 "source": "app_file"216 },217 {218 "quest": "Off the Grid",219 "confidence": 0.93,220 "evidence": "local transformers load, no openai/anthropic/gemini imports",221 "source": "app_file"222 },223 {224 "quest": "Thousand Token Wood",225 "confidence": 0.95,226 "evidence": "joyful Gradio toy... impossible weather... miniature stage plays",227 "source": "readme"228 },229 {230 "quest": "Sharing is Caring",231 "confidence": 0.85,232 "evidence": "one-click download pack, downloadable postcard image, Share Card",233 "source": "readme"234 }235 ],236 "signal_profile": "both",237 "audit": "Kept Well-Tuned (published LoRA repo + PeftModel import), Tiny Titan (SmolLM2-135M, under 4B), Off the Grid (local transformers, no cloud API imports), Thousand Token Wood (joyful stage-play framing), Sharing is Caring (download pack/postcard/Share Card). Dropped Off-Brand: cited evidence is generated media outputs, not custom frontend; no gr.HTML/Blocks(css=) verifiable in APP_FILE. Profile both."238 },239 {240 "slug": "Trollsona",241 "matches": [242 {243 "quest": "Thousand Token Wood",244 "confidence": 0.97,245 "evidence": "Track: An Adventure in Thousand Token Wood; playful troll alter ego; goblin meter",246 "source": "readme"247 },248 {249 "quest": "Tiny Titan",250 "confidence": 0.93,251 "evidence": "compact model derived from Qwen/Qwen2.5-3B-Instruct; lightweight Qwen 0.5B fallback",252 "source": "readme"253 },254 {255 "quest": "Off the Grid",256 "confidence": 0.85,257 "evidence": "Local Hugging Face Transformers generation path for the primary AI runtime",258 "source": "readme"259 }260 ],261 "signal_profile": "both",262 "audit": "KEPT Thousand Token Wood (track+goblin meter), Tiny Titan (3B/0.5B Qwen <=4B), Off the Grid (local transformers, no cloud API; conf lowered as model off-by-default). DROPPED Well-Tuned: derived/compact/qkvfp16=quantization not fine-tune, no LoRA/PEFT/SFT. DROPPED Field Notes: features/fallback list is ordinary README. APP_FILE corroborates both axes."263 },264 {265 "slug": "gitopadesh",266 "matches": [267 {268 "quest": "Thousand Token Wood",269 "confidence": 0.9,270 "evidence": "Lord Krishna himself responds in first person citing exact Chapter and Verse",271 "source": "readme"272 },273 {274 "quest": "Sharing is Caring",275 "confidence": 0.8,276 "evidence": "Your Shloka Card, Download and Share",277 "source": "app_file"278 },279 {280 "quest": "Off-Brand",281 "confidence": 0.6,282 "evidence": "multilingual TRANSLATIONS UI, Battlefield Map, PIL shloka card",283 "source": "app_file"284 }285 ],286 "signal_profile": "both",287 "audit": "KEPT Thousand Token Wood (Krishna first-person verse persona). KEPT Sharing (Shloka Card download string plus PIL card). KEPT Off-Brand at 0.6, evidence retied to findable strings. DROPPED Field Notes (prompt blurb, not a devlog). DROPPED Backyard AI (targets every seeker, generic not named). No Tiny Titan (Qwen 7B); no Off the Grid (HF hosted, not local)."288 },289 {290 "slug": "roadb-other-screen",291 "matches": [292 {293 "quest": "Llama Champion",294 "confidence": 0.97,295 "evidence": "from llama_cpp import Llama",296 "source": "app_file"297 },298 {299 "quest": "Off the Grid",300 "confidence": 0.85,301 "evidence": "Qwen GGUF through llama.cpp via llama-cpp-python",302 "source": "app_file"303 },304 {305 "quest": "Thousand Token Wood",306 "confidence": 0.93,307 "evidence": "interactive fiction experience... cinematic Other Screen... strange mirror",308 "source": "readme"309 },310 {311 "quest": "Off-Brand",312 "confidence": 0.9,313 "evidence": "from gradio import Server; custom frontend: index.html",314 "source": "app_file"315 },316 {317 "quest": "Sharing is Caring",318 "confidence": 0.88,319 "evidence": "download a souvenir card... export a synthetic-style trace",320 "source": "readme"321 }322 ],323 "signal_profile": "both",324 "audit": "KEPT all 5; evidence verified in source. Off the Grid conf trimmed: GGUF/llama-cpp local, no cloud API in visible app_file (truncated). TTW evidence tightened to findable README quote. NOT added Tiny Titan (Qwen3.5-9B = 9B, over 4B) or Modal (tag only, absent from README/app_file text)."325 },326 {327 "slug": "dreamwall-mc",328 "matches": [329 {330 "quest": "Thousand Token Wood",331 "confidence": 0.92,332 "evidence": "hatch a NeuroPet from a prompt, carve creature's memory",333 "source": "readme"334 },335 {336 "quest": "Sharing is Caring",337 "confidence": 0.45,338 "evidence": "every prompt becomes part of a public server museum",339 "source": "readme"340 }341 ],342 "signal_profile": "readme_only",343 "audit": "Kept Thousand Token Wood (playful Minecraft creature/art game). Kept Sharing is Caring at 0.45 (public museum of artifacts, but no export/push_to_hub in truncated APP_FILE). Dropped Off the Grid: no local model load (MODEL_ID is a hashlib fingerprint, declared_models empty); rested on API absence plus hedged README. Dropped Field Notes: \"Why This Is Different\" is positioning, not a build write-up."344 },345 {346 "slug": "persona-atlas",347 "matches": [348 {349 "quest": "Best Agent",350 "confidence": 0.93,351 "evidence": "agent runs web searches, pulls portrait, assembles profile",352 "source": "readme"353 },354 {355 "quest": "Thousand Token Wood",356 "confidence": 0.85,357 "evidence": "Put Socrates, Churchill, and Sam Altman in the same room",358 "source": "readme"359 }360 ],361 "signal_profile": "both",362 "audit": "KEPT Best Agent: real autonomous loop (web search, assemble dossier); DDGS confirms. KEPT Thousand Token Wood: playful 'same room' AI experiment. DROPPED Sharing is Caring: portrait stored with run is internal, not user export. DROPPED Off-Brand: matplotlib/UMAP/html.escape are generic, not custom CSS/gr.HTML; only plain import gradio. Not added: Off the Grid (remote HF novita), Tiny Titan (26B)."363 },364 {365 "slug": "ContextForge",366 "matches": [367 {368 "quest": "Off the Grid",369 "confidence": 0.85,370 "evidence": "from transformers import AutoModelForCausalLM, AutoTokenizer; no cloud imports",371 "source": "app_file"372 },373 {374 "quest": "Tiny Titan",375 "confidence": 0.85,376 "evidence": "DEFAULT_MODEL_ID = \"Qwen/Qwen2.5-0.5B-Instruct\"",377 "source": "app_file"378 },379 {380 "quest": "Backyard AI",381 "confidence": 0.68,382 "evidence": "Backyard AI Fit - Built for real builders using AI coding agents",383 "source": "readme"384 }385 ],386 "signal_profile": "both",387 "audit": "KEPT Off the Grid: transformers/Qwen load, no cloud imports. KEPT Tiny Titan: default Qwen2.5-0.5B; 32B is optional high tier. KEPT Backyard AI, conf lowered (builder persona, not one named person). DROPPED Best Agent: AGENTIC_LOOP is a prompt-template tag and the seven modules are a fixed sequential pipeline, not autonomous tool-use; evidence wrongly attributed to readme."388 },389 {390 "slug": "lovegpt",391 "matches": [392 {393 "quest": "Llama Champion",394 "confidence": 0.78,395 "evidence": "llama-server -hf ysn-rfd/Phi-4-mini-instruct-GGUF:Q4_0",396 "source": "readme"397 },398 {399 "quest": "Tiny Titan",400 "confidence": 0.75,401 "evidence": "local Phi-4-mini-instruct GGUF runtime",402 "source": "readme"403 }404 ],405 "signal_profile": "readme_only",406 "audit": "Fields truncated; judged visible text. KEPT Llama Champion (README llama-server pulls Phi-4-mini GGUF; lower conf, runtime is optional). KEPT Tiny Titan (Phi-4-mini ~3.8B). DROPPED Off-Brand: Flutter separate, Space is default Gradio adapter, no gr.HTML/css= visible. DROPPED Well-Tuned: vague training-pipeline line only. DROPPED Sharing: encrypted key-locked JSONL, not shareable."407 },408 {409 "slug": "her",410 "matches": [411 {412 "quest": "Off-Brand",413 "confidence": 0.95,414 "evidence": "Gradio Server mode, FastAPI server, React app calls custom routes",415 "source": "app_file"416 }417 ],418 "signal_profile": "app_only",419 "audit": "KEPT Off-Brand: APP_FILE shows Gradio Server mode, FastAPI server, React app on bespoke /api routes; tightened evidence. DROPPED Backyard AI: README is a general dev tool for coding-agent sessions, not a specific named person the quest needs. No new matches: deterministic analyzer not agentic; no clear export or local-model load in truncated text. profile -> app_only."420 },421 {422 "slug": "ux-crime-scene",423 "matches": [424 {425 "quest": "Modal",426 "confidence": 0.97,427 "evidence": "Qwen2.5-VL-32B-Instruct on Modal (vLLM, A100-80GB)",428 "source": "readme"429 },430 {431 "quest": "Off-Brand",432 "confidence": 0.9,433 "evidence": "STATIC_CSS reads styles.css; --asset CSS var overrides",434 "source": "app_file"435 },436 {437 "quest": "Thousand Token Wood",438 "confidence": 0.92,439 "evidence": "hard-boiled, film-noir detective; detective thriller",440 "source": "readme"441 },442 {443 "quest": "Sharing is Caring",444 "confidence": 0.85,445 "evidence": "every investigation gets a unique, shareable link",446 "source": "readme"447 },448 {449 "quest": "Best Agent",450 "confidence": 0.78,451 "evidence": "imports investigate_agentic from detective",452 "source": "app_file"453 }454 ],455 "signal_profile": "both",456 "audit": "KEPT all 5; cited segments verified present. Off-Brand trimmed to 0.90 (real STATIC_CSS + base64 --asset CSS overrides). Sharing 0.85. Best Agent 0.78 (only investigate_agentic import visible, loop body off-segment). DROPPED: Tiny Titan (32B not <=4B), Off the Grid (remote Modal vLLM, not local), Field Notes (truncated tech table, no write-up)."457 },458 {459 "slug": "Tone-Bridge",460 "matches": [461 {462 "quest": "Off-Brand",463 "confidence": 0.8,464 "evidence": "app = gr.Server()",465 "source": "app_file"466 },467 {468 "quest": "Off the Grid",469 "confidence": 0.85,470 "evidence": "AutoModelForCausalLM from transformers; no cloud API imports",471 "source": "app_file"472 },473 {474 "quest": "Field Notes",475 "confidence": 0.6,476 "evidence": "README has Problem / What it does / How it works write-up",477 "source": "readme"478 }479 ],480 "signal_profile": "both",481 "audit": "KEPT Off-Brand (gr.Server() in app_file; 0.95->0.8). KEPT Off the Grid (local Qwen, 4bit, no cloud API; ->0.85). KEPT Field Notes (Problem/What/How write-up; ->0.6). DROPPED Backyard AI: broad learner audience, not a named person. DROPPED Tiny Titan: core model is Qwen3-14B; 0.6B is only auxiliary TTS."482 },483 {484 "slug": "family-care-network",485 "matches": [486 {487 "quest": "Backyard AI",488 "confidence": 0.96,489 "evidence": "Specific real user: a Ghanaian family coordinator checking on elders",490 "source": "readme"491 },492 {493 "quest": "Well-Tuned",494 "confidence": 0.9,495 "evidence": "\"Adwuma Pa Akan Whisper fine-tune\", \"fine_tuned\" ASR choice",496 "source": "app_file"497 },498 {499 "quest": "Field Notes",500 "confidence": 0.7,501 "evidence": "See CODEX_BUILD_LOG.md and HACKATHON_TODO.md documenting build",502 "source": "readme"503 }504 ],505 "signal_profile": "both",506 "audit": "KEPT Backyard AI, Well-Tuned (literal fine_tuned ASR choice), Field Notes (build-log docs). DROPPED Modal: app imports local services.modal_client, not import modal; README never cites Modal. DROPPED Best Agent: \"agentic\" is marketing for rule-based relay, no tool/function-call loop. DROPPED Off-Brand: \"custom Gradio UI\" is only a badge target; app is default gr tables."507 },508 {509 "slug": "Case-Lantern",510 "matches": [511 {512 "quest": "Llama Champion",513 "confidence": 0.99,514 "evidence": "from llama_cpp import Llama",515 "source": "app_file"516 },517 {518 "quest": "Off the Grid",519 "confidence": 0.95,520 "evidence": "Backend: llama-cpp-python (GGUF, runs on free CPU Spaces); no cloud API",521 "source": "app_file"522 },523 {524 "quest": "Thousand Token Wood",525 "confidence": 0.92,526 "evidence": "fictional medical mystery game; Thousand Token Wood track",527 "source": "readme"528 },529 {530 "quest": "Off-Brand",531 "confidence": 0.85,532 "evidence": "Dark glassmorphism theme, custom CSS",533 "source": "readme"534 },535 {536 "quest": "Well-Tuned",537 "confidence": 0.7,538 "evidence": "Uses fine-tuned model published on HF (Qwen3.5-Medical-GSPO)",539 "source": "readme"540 }541 ],542 "signal_profile": "both",543 "audit": "KEPT Llama Champion (literal import) and Off the Grid (app docstring llama-cpp-python GGUF, no cloud client). KEPT Thousand Token Wood, Off-Brand, Well-Tuned from README. DROPPED Tiny Titan: README says ~4.66B params, exceeds the <=4B threshold (only constraint cited is under 32B). No missed quests: no tool-calling, no export artifact, README is a badge sheet not a devlog."544 },545 {546 "slug": "planpalette",547 "matches": [548 {549 "quest": "Off the Grid",550 "confidence": 0.72,551 "evidence": "diffusers AutoPipelineForText2Image; no proprietary API imports",552 "source": "app_file"553 }554 ],555 "signal_profile": "app_only",556 "audit": "KEPT Off the Grid: diffusers AutoPipelineForText2Image in APP_FILE, open-weight SDXL run locally, no cloud API. DROPPED Backyard AI: 'visualization artists' is a generic profession, no named person. DROPPED Field Notes: README is a feature doc, not a build write-up. DROPPED Sharing is Caring: 'Final PNG output' is a displayed render, no export button. Only app_file yields a signal."557 },558 {559 "slug": "cloud-parade-cabinet",560 "matches": [561 {562 "quest": "Thousand Token Wood",563 "confidence": 0.97,564 "evidence": "kinetic Gradio toy where a hosted small model invents a tiny impossible parade",565 "source": "readme"566 },567 {568 "quest": "OpenBMB",569 "confidence": 0.97,570 "evidence": "OPENBMB_MODEL = ... openbmb/MiniCPM4-8B",571 "source": "app_file"572 },573 {574 "quest": "Nemotron",575 "confidence": 0.97,576 "evidence": "NVIDIA_MODEL = ... nvidia/llama-3.1-nemotron-nano-8b-v1",577 "source": "app_file"578 }579 ],580 "signal_profile": "both",581 "audit": "KEPT Thousand Token Wood (toy/art framing), OpenBMB (openbmb/MiniCPM4-8B), Nemotron (nvidia nemotron-nano-8b). DROPPED Sharing is Caring: only script filenames cited, no download/export or push_to_hub shown. Off the Grid absent: readme 'Cloud API allowed', app uses InferenceClient+NVIDIA. Tiny Titan absent: 7B/8B. Modal tag has no import modal in text."582 },583 {584 "slug": "neilA",585 "matches": [586 {587 "quest": "Thousand Token Wood",588 "confidence": 0.9,589 "evidence": "that \"it finally understood me\" moment is the payoff",590 "source": "readme"591 },592 {593 "quest": "Off the Grid",594 "confidence": 0.65,595 "evidence": "LocalBrain/make_brain; model loaded at MODULE level, no proprietary API imports",596 "source": "app_file"597 },598 {599 "quest": "Off-Brand",600 "confidence": 0.55,601 "evidence": "css/theme on Blocks() so custom styling applies whichever version",602 "source": "app_file"603 }604 ],605 "signal_profile": "both",606 "audit": "Kept Thousand Token Wood (verbatim payoff quote; playful alien game). Kept Off the Grid, tightened to LocalBrain/make_brain at module level, no openai/anthropic imports; lowered to 0.65 (brain/model truncated). Kept Off-Brand at 0.55: only css/theme-on-Blocks() comment visible, no concrete css= arg. DROPPED Field Notes: Architecture+SPEC link is product docs, not a build write-up/devlog."607 },608 {609 "slug": "First-Principle-AI",610 "matches": [611 {612 "quest": "Llama Champion",613 "confidence": 0.96,614 "evidence": "GGUF model through the official llama.cpp llama-server release",615 "source": "readme"616 },617 {618 "quest": "Off the Grid",619 "confidence": 0.8,620 "evidence": "hf_hub_download of model-Q8_0.gguf, run via local llama-server, no cloud client",621 "source": "app_file"622 },623 {624 "quest": "Field Notes",625 "confidence": 0.55,626 "evidence": "Runtime Notes: model repo, file, runtime, hardware target, llama.cpp settings",627 "source": "readme"628 }629 ],630 "signal_profile": "both",631 "audit": "Both fields truncated; audited visible text only. KEPT Llama Champion (README: GGUF via official llama.cpp llama-server); fixed source to readme. FIXED Off the Grid: cited phrase was README; app_file shows hf_hub_download of model-Q8_0.gguf, no cloud client, kept on app_file at 0.8. ADDED Field Notes (Runtime Notes). No size named, no Tiny Titan; no custom CSS, no Off-Brand."632 },633 {634 "slug": "legawa",635 "matches": [636 {637 "quest": "Best Agent",638 "confidence": 0.95,639 "evidence": "Asisten multi-agen ... analis_ruu, peneliti, penyusun, surat",640 "source": "readme"641 }642 ],643 "signal_profile": "both",644 "audit": "KEPT Best Agent: README multi-agen + four named agents; APP_FILE imports analis_ruu/peneliti/penyusun/surat with tools + query-expansion search. DROPPED Backyard AI: targets legislators as a profession, not one named person. DROPPED Off the Grid: default is HF Inference API (cloud); llama.cpp only optional. No Tiny Titan (9B/30B/8B all >4B)."645 },646 {647 "slug": "come-and-compare",648 "matches": [],649 "signal_profile": "none",650 "audit": "Dropped both. Backyard AI: README targets \"millions of shoppers in India,\" mass-market not one named person. Field Notes: those are pitch headers, not a build write-up. No new match: Tiny Titan ineligible (Qwen2.5-7B); Off the Grid fails (HF InferenceClient API, no local load); Best Agent/Off-Brand unevidenced in truncated text."651 },652 {653 "slug": "hackathon-advisor",654 "matches": [655 {656 "quest": "Well-Tuned",657 "confidence": 0.97,658 "evidence": "published PEFT LoRA adapter for MiniCPM5",659 "source": "readme"660 },661 {662 "quest": "Llama Champion",663 "confidence": 0.95,664 "evidence": "runtime query embeddings computed through llama.cpp",665 "source": "readme"666 },667 {668 "quest": "Nemotron",669 "confidence": 0.95,670 "evidence": "Nemotron Speech Streaming via NVIDIA NeMo ASR",671 "source": "readme"672 },673 {674 "quest": "Best Agent",675 "confidence": 0.9,676 "evidence": "AdvisorEngine tool-call planning, resolve_tool_call, tool_schemas",677 "source": "app_file"678 },679 {680 "quest": "Off the Grid",681 "confidence": 0.92,682 "evidence": "local MiniCPM5 transformers plus llama.cpp GGUF, no cloud API",683 "source": "app_file"684 },685 {686 "quest": "Tiny Titan",687 "confidence": 0.95,688 "evidence": "runs openbmb/MiniCPM5-1B, a 1B model",689 "source": "readme"690 }691 ],692 "signal_profile": "both",693 "audit": "All 6 proposals verified as genuine. KEPT Well-Tuned, Llama Champion, Nemotron, Best Agent. Added strong misses Off the Grid (fully local; only openai ref is local whisper, not cloud API) and Tiny Titan (MiniCPM5-1B). DROPPED Off-Brand and Modal only for the 6-item cap; both still valid. OpenBMB also valid but cut by cap."694 },695 {696 "slug": "wpl-discovery",697 "matches": [698 {699 "quest": "Off-Brand",700 "confidence": 0.55,701 "evidence": "Gradio 6 with Worcestershire County Council brand colours",702 "source": "readme"703 }704 ],705 "signal_profile": "readme_only",706 "audit": "Kept Off-Brand: README cites custom council brand colours over default Gradio. Dropped Backyard AI: track label, serves a whole county not a named person. Dropped Best Agent: query_tool is keyword/RAG routing, not tool-calling/loop. Dropped Field Notes: brief how-it-works blurb, no real write-up. No Off the Grid (Anthropic+HF cloud), no Tiny Titan (Qwen 32B)."707 },708 {709 "slug": "The-Shrine",710 "matches": [711 {712 "quest": "Thousand Token Wood",713 "confidence": 0.95,714 "evidence": "Build something delightful that wouldn't exist without AI",715 "source": "readme"716 },717 {718 "quest": "Off-Brand",719 "confidence": 0.85,720 "evidence": "Custom Canvas + vanilla JS (60+ monologue phrases, 5 phases)",721 "source": "readme"722 }723 ],724 "signal_profile": "readme_only",725 "audit": "Kept Thousand Token Wood (README states track verbatim; delightful AI art). Kept Off-Brand (README: Custom Canvas + vanilla JS); lowered conf since canvas code absent from truncated app.py. Denied Off the Grid: README self-claims it but APP_FILE calls DashScope + OpenRouter cloud APIs. No Well-Tuned (badge-only) or Tiny Titan (Qwen2.5-7B)."726 },727 {728 "slug": "octopus-ai",729 "matches": [730 {731 "quest": "Off-Brand",732 "confidence": 0.9,733 "evidence": "mission-control dark palette, gr.HTML blocks, CSS-only animations",734 "source": "app_file"735 },736 {737 "quest": "Well-Tuned",738 "confidence": 0.55,739 "evidence": "4 specialized arms trained with Structurally Adaptive Learning (SAL)",740 "source": "readme"741 },742 {743 "quest": "Best Agent",744 "confidence": 0.45,745 "evidence": "structural brain and arms detect damage, reroute, and recover",746 "source": "readme"747 }748 ],749 "signal_profile": "both",750 "audit": "Kept Off-Brand: real custom dark palette, gr.HTML panels, CSS animations in APP_FILE. Kept Well-Tuned: README claims arms trained via SAL; conf lowered (no script). Kept Best Agent, tightened/lowered: brain routes arms, reroute/recover. DROPPED Thousand Token Wood: engineering demo, not art/whimsy. No Off the Grid: 7B open-weight but no local-load signal. No cloud API."751 },752 {753 "slug": "VoiceGate",754 "matches": [755 {756 "quest": "Off the Grid",757 "confidence": 0.8,758 "evidence": "all models loaded from local COMFY_DIR paths; ComfyUI on 127.0.0.1; no cloud API client",759 "source": "app_file"760 },761 {762 "quest": "Tiny Titan",763 "confidence": 0.88,764 "evidence": "Qwen3-ASR-1.7B and Qwen3-ForcedAligner-0.6B model paths",765 "source": "app_file"766 }767 ],768 "signal_profile": "app_only",769 "audit": "Kept Off the Grid: models all from local COMFY_DIR paths, ComfyUI on 127.0.0.1, no openai/anthropic/gemini/cohere import; tightened evidence, confidence 0.8. Kept Tiny Titan: Qwen3-ASR-1.7B and 0.6B appear verbatim, both <=4B. Dropped potential OpenBMB: voxcpm/VoxCPM2 path present but no openbmb/ namespace or MiniCPM signal. No Off-Brand/Agent/Backyard evidence (UI truncated; generic README)."770 },771 {772 "slug": "AI-Puppet-Theater",773 "matches": [774 {775 "quest": "OpenBMB",776 "confidence": 0.97,777 "evidence": "OpenBMB model id: openbmb/MiniCPM5-1B",778 "source": "app_file"779 },780 {781 "quest": "Tiny Titan",782 "confidence": 0.95,783 "evidence": "openbmb/MiniCPM5-1B names a 1B model",784 "source": "app_file"785 },786 {787 "quest": "Off-Brand",788 "confidence": 0.94,789 "evidence": "CUSTOM_CSS overrides .gradio-container background",790 "source": "app_file"791 },792 {793 "quest": "Thousand Token Wood",794 "confidence": 0.9,795 "evidence": "short interactive puppet shows from a user premise",796 "source": "readme"797 },798 {799 "quest": "Off the Grid",800 "confidence": 0.6,801 "evidence": "BACKEND_CHOICES deterministic/openbmb, no cloud imports",802 "source": "app_file"803 }804 ],805 "signal_profile": "both",806 "audit": "Kept OpenBMB, Tiny Titan, Off-Brand, Thousand Token Wood, Off the Grid; all evidence present in cited segments. Dropped Best Agent: named functions are imported show-builder helpers, no visible tool-calling, planner, or multi-step loop in truncated source. Lowered Off the Grid to 0.6 (model unloaded, deterministic fallback, no cloud client visible). signal_profile stays both."807 },808 {809 "slug": "LocalDuo",810 "matches": [811 {812 "quest": "Tiny Titan",813 "confidence": 0.93,814 "evidence": "model_id = \"Qwen/Qwen3.5-2B\"",815 "source": "app_file"816 },817 {818 "quest": "Off the Grid",819 "confidence": 0.8,820 "evidence": "AutoModelForImageTextToText loads Qwen via transformers; no cloud API imports",821 "source": "app_file"822 }823 ],824 "signal_profile": "app_only",825 "audit": "Read full 1903-char APP_FILE. KEPT Tiny Titan: active model_id Qwen/Qwen3.5-2B is 2B (9B commented out). KEPT Off the Grid: local transformers load, no openai/anthropic/genai/cohere imports. DROPPED Best Agent: code only imports playwright/BeautifulSoup/TTS plus stop flags; scripted browsing is not function-calling or a planner loop. README is templated HF stub, so app_only."826 },827 {828 "slug": "PocketWorld-Studio",829 "matches": [830 {831 "quest": "Thousand Token Wood",832 "confidence": 0.62,833 "evidence": "WORLD_THEMES cozy_fantasy, sci_fi_station, haunted_mystery with npc/item/landmark keys",834 "source": "app_file"835 }836 ],837 "signal_profile": "app_only",838 "audit": "KEPT Thousand Token Wood (tightened evidence, 0.75->0.62): playful multi-world studio with NPCs/items/landmarks is in APP_FILE, but no AI/LLM visible (static Kenney CC0 tile renderer, declared_models empty). README is HF boilerplate. Did NOT add Off-Brand (no gr.HTML/css visible) or Sharing (no download/gr.File). No model signals for other quests."839 },840 {841 "slug": "voice-sales-logger",842 "matches": [843 {844 "quest": "Backyard AI",845 "confidence": 0.9,846 "evidence": "Voice-driven sales logger for a produce vendor",847 "source": "readme"848 },849 {850 "quest": "Modal",851 "confidence": 0.8,852 "evidence": "if not ON_SPACE: import modal",853 "source": "app_file"854 },855 {856 "quest": "Off the Grid",857 "confidence": 0.58,858 "evidence": "transformers AutoModelForCausalLM + nemo.collections.asr; no cloud API",859 "source": "app_file"860 },861 {862 "quest": "Off-Brand",863 "confidence": 0.45,864 "evidence": "from gradio_wizardcapture import WizardCapture (vendored component)",865 "source": "app_file"866 }867 ],868 "signal_profile": "both",869 "audit": "KEPT Backyard AI: README verbatim, real produce-vendor task (0.92>0.9). KEPT Modal: import present but gated by ON_SPACE (0.88>0.8). KEPT Off the Grid: transformers+NeMo loaded, no openai/anthropic/genai/cohere visible; trimmed 0.65>0.58 (APP_FILE truncated). ADDED Off-Brand low-conf: vendored WizardCapture component. DROPPED Nemotron candidate: generic NeMo RNNT, no named nvidia/Parakeet repo."870 },871 {872 "slug": "wan2-2-fp8da-aoti-14B-fast",873 "matches": [874 {875 "quest": "Off the Grid",876 "confidence": 0.75,877 "evidence": "diffusers WanImageToVideoPipeline + torchao loaded locally, no cloud API imports",878 "source": "app_file"879 },880 {881 "quest": "Off-Brand",882 "confidence": 0.6,883 "evidence": "get_timestamp_js custom JS querySelector('#generated-video video')",884 "source": "app_file"885 }886 ],887 "signal_profile": "app_only",888 "audit": "KEPT Off the Grid: APP_FILE loads diffusers Wan i2v pipeline + torchao locally; no openai/anthropic/genai/cohere visible. Lowered 0.82->0.75 (file truncated). KEPT Off-Brand: custom get_timestamp_js JS present, beyond default Gradio. README templated HF link, no Field Notes. 'mcp-server' tag alone weak for Best Agent. Empty models + 14B title => no Tiny Titan."889 },890 {891 "slug": "nutrilens",892 "matches": [893 {894 "quest": "Off-Brand",895 "confidence": 0.92,896 "evidence": "CUSTOM_CSS with .nutrilens-report, .summary-card, .tip-card classes",897 "source": "app_file"898 },899 {900 "quest": "Best Agent",901 "confidence": 0.72,902 "evidence": "identifies each ingredient, looks up USDA, finds PubMed studies",903 "source": "readme"904 }905 ],906 "signal_profile": "both",907 "audit": "KEEP Off-Brand: real CUSTOM_CSS classes in app_file. KEEP Best Agent (0.82->0.72): multi-step tools (VLM, USDA, PubMed lookups), fixed pipeline. DROP Backyard AI: targets \"anyone,\" no named person. DROP Field Notes: motivation/usage only, no build write-up. Not Off the Grid (HF cloud client) nor Tiny Titan (Qwen3.6-27B)."908 },909 {910 "slug": "business-order-assistant",911 "matches": [912 {913 "quest": "Modal",914 "confidence": 0.9,915 "evidence": "defaults to ...chat-query.modal.run endpoints called via requests",916 "source": "app_file"917 },918 {919 "quest": "Off-Brand",920 "confidence": 0.95,921 "evidence": "CUSTOM_CSS terminal/amber, JetBrains Mono, #app-title and #sidebar",922 "source": "app_file"923 }924 ],925 "signal_profile": "app_only",926 "audit": "README is templated boilerplate, so app_only stands. KEPT Modal: real *.modal.run endpoints invoked via requests; no `import modal`, lowered to 0.90. KEPT Off-Brand: genuine CUSTOM_CSS, JetBrains Mono, amber/dark, bespoke #app-title/#sidebar. Tightened both evidence quotes. Skipped Tiny Titan (Qwen2.5-7B >4B), Off the Grid (model on Modal, not local), Field Notes, Best Agent."927 },928 {929 "slug": "Kasualdad_LFED",930 "matches": [931 {932 "quest": "Off the Grid",933 "confidence": 0.95,934 "evidence": "All inference via llama.cpp + local GGUF. No API calls. No cloud.",935 "source": "readme"936 },937 {938 "quest": "Llama Champion",939 "confidence": 0.95,940 "evidence": "llama.cpp as the sole inference backend. Q4_K_M quantization.",941 "source": "readme"942 },943 {944 "quest": "Well-Tuned",945 "confidence": 0.93,946 "evidence": "Fine-tuned Qwen2.5-Coder-7B on 1,200+ synthetic NL→SQL pairs via Unsloth QLoRA",947 "source": "readme"948 },949 {950 "quest": "Off-Brand",951 "confidence": 0.75,952 "evidence": "Custom design system (Linear/Vercel inspired), WCAG AA, Inter + JetBrains Mono",953 "source": "readme"954 },955 {956 "quest": "Backyard AI",957 "confidence": 0.85,958 "evidence": "school district admin (principal, superintendent, department head) types a question",959 "source": "readme"960 },961 {962 "quest": "Modal",963 "confidence": 0.7,964 "evidence": "Fine-tuned ... via Unsloth QLoRA on Modal A10G",965 "source": "readme"966 }967 ],968 "signal_profile": "both",969 "audit": "All 6 quotes verified in README badge table / What It Does. APP_FILE confirms llama.cpp (docstring), DuckDB, gradio+spaces; NO openai/anthropic/google/cohere, so Off the Grid stands. Tiny Titan correctly absent (7B). Lowered Off-Brand to 0.75 (badge self-claim, no CSS in truncated app.py) and Modal to 0.70 (single training mention)."970 },971 {972 "slug": "thousand-token-wood-sim",973 "matches": [974 {975 "quest": "Thousand Token Wood",976 "confidence": 0.99,977 "evidence": "Five woodland creatures... trade goods for pebbles, gossip, scheme, and panic",978 "source": "readme"979 },980 {981 "quest": "Well-Tuned",982 "confidence": 0.97,983 "evidence": "ttw-trader-0.5b (fine-tuned)... distilled from a 3B teacher",984 "source": "readme"985 },986 {987 "quest": "OpenBMB",988 "confidence": 0.96,989 "evidence": "Bramble (squirrel) | MiniCPM3-4B | OpenBMB",990 "source": "readme"991 },992 {993 "quest": "Nemotron",994 "confidence": 0.96,995 "evidence": "Fenn (fox) | Nemotron-Mini-4B | NVIDIA",996 "source": "readme"997 },998 {999 "quest": "Modal",1000 "confidence": 0.95,1001 "evidence": "served with vLLM on Modal; modal deploy serve.py",1002 "source": "readme"1003 },1004 {1005 "quest": "Field Notes",1006 "confidence": 0.9,1007 "evidence": "Field notes: Hugging Face article | Medium",1008 "source": "readme"1009 }1010 ],1011 "signal_profile": "both",1012 "audit": "All 6 first-pass matches verified; KEPT 5 (TTW, Well-Tuned, OpenBMB, Nemotron, Field Notes) with tightened quotes. DROPPED Sharing is Caring (valid but lower-value) to ADD Modal, clearer and also in APP_FILE (ModalLLM, modal deploy serve.py). Off the Grid not awarded: Modal cloud, not local. Tiny Titan plausible but cast has gpt-oss-20b; capped at 6."1013 },1014 {1015 "slug": "memrl-canvas",1016 "matches": [1017 {1018 "quest": "Off the Grid",1019 "confidence": 0.76,1020 "evidence": "transformers AutoModelForCausalLM load, no cloud API client",1021 "source": "app_file"1022 },1023 {1024 "quest": "Off-Brand",1025 "confidence": 0.5,1026 "evidence": "mounts StaticFiles and FileResponse for a custom frontend",1027 "source": "app_file"1028 }1029 ],1030 "signal_profile": "app_only",1031 "audit": "KEEP Off the Grid: transformers AutoModelForCausalLM present, no cloud client visible; conf trimmed (no model name, truncated). FIX Off-Brand: bare gradio Server import weak; re-anchored to StaticFiles/FileResponse frontend. DROP Best Agent: episodic_memory/q_value is a seeded SQLite lookup, no loop/tool-call visible, so RL agent loop is speculative. README is HF boilerplate."1032 },1033 {1034 "slug": "roast-my-repo",1035 "matches": [1036 {1037 "quest": "Off-Brand",1038 "confidence": 0.93,1039 "evidence": "custom terminal CSS, JetBrains Mono, scanline overlay effect",1040 "source": "app_file"1041 },1042 {1043 "quest": "OpenBMB",1044 "confidence": 0.95,1045 "evidence": "MiniCPM4-8B (openbmb/MiniCPM4-8B) (OpenBMB)",1046 "source": "readme"1047 },1048 {1049 "quest": "Modal",1050 "confidence": 0.95,1051 "evidence": "MiniCPM4-8B via vLLM on Modal, Modal A10G GPU",1052 "source": "readme"1053 },1054 {1055 "quest": "Backyard AI",1056 "confidence": 0.7,1057 "evidence": "final-year CS students wanting honest feedback on GitHub portfolio before jobs",1058 "source": "readme"1059 },1060 {1061 "quest": "Thousand Token Wood",1062 "confidence": 0.55,1063 "evidence": "brutal, funny critique; tells you what your friends won't",1064 "source": "readme"1065 },1066 {1067 "quest": "Sharing is Caring",1068 "confidence": 0.5,1069 "evidence": "Generated README you can copy and use immediately",1070 "source": "readme"1071 }1072 ],1073 "signal_profile": "both",1074 "audit": "All 5 verified. KEEP Off-Brand (app_file CSS present), OpenBMB, Modal (exact readme). Backyard AI lowered to 0.7 (broad persona, not one named person). Thousand Token Wood kept 0.55 (roast whimsy real but core is a utility). ADDED Sharing is Caring 0.5 (copyable generated README). Confirmed NO Tiny Titan (8B) and NO Off the Grid (Modal cloud + OpenAI endpoint + Groq)."1075 },1076 {1077 "slug": "tiny-dispatch-coach",1078 "matches": [],1079 "signal_profile": "none",1080 "audit": "Dropped Backyard AI: label-echo for a generic segment (delivery teams), not a named person. Off the Grid not awarded: shipped app is a deterministic offline planner with NO model loaded; MiniCPM/Llama-via-llama.cpp is only a future swap, declared_models empty, no runtime/cloud API. Llama Champion/Tiny Titan/OpenBMB rest on that conditional mention. No frontend/agent/finetune signals."1081 },1082 {1083 "slug": "dm-order-desk",1084 "matches": [1085 {1086 "quest": "Off the Grid",1087 "confidence": 0.9,1088 "evidence": "AutoModelForCausalLM.from_pretrained(MODEL_ID), local transformers",1089 "source": "app_file"1090 },1091 {1092 "quest": "Tiny Titan",1093 "confidence": 0.97,1094 "evidence": "Model: Qwen/Qwen2.5-1.5B-Instruct, about 1.5B",1095 "source": "readme"1096 },1097 {1098 "quest": "Backyard AI",1099 "confidence": 0.62,1100 "evidence": "designed for home bakers, farmers market vendors, WhatsApp sellers",1101 "source": "readme"1102 }1103 ],1104 "signal_profile": "both",1105 "audit": "Both fields truncated. KEPT Off the Grid: local transformers/torch loads Qwen2.5-1.5B, no cloud client visible; trimmed conf. KEPT Tiny Titan: README names Qwen2.5-1.5B (~1.5B), verbatim. KEPT Backyard AI at 0.62: targets persona categories not one named person (Maya/Sam are customers). No fine-tune, custom UI, export, llama.cpp, Modal, OpenBMB, Nemotron, or agent loop evident."1106 },1107 {1108 "slug": "neighbourhood-guide",1109 "matches": [1110 {1111 "quest": "Llama Champion",1112 "confidence": 0.95,1113 "evidence": "from llama_cpp import Llama; LLAMA_CPP_FILE *Q4_K_M.gguf",1114 "source": "app_file"1115 },1116 {1117 "quest": "Backyard AI",1118 "confidence": 0.85,1119 "evidence": "Backyard AI track: solve a real problem for someone you know",1120 "source": "readme"1121 },1122 {1123 "quest": "Tiny Titan",1124 "confidence": 0.85,1125 "evidence": "Tiny Aya 3.35B, MagpieTTS 357M, Cohere Transcribe 2B",1126 "source": "readme"1127 },1128 {1129 "quest": "Off the Grid",1130 "confidence": 0.75,1131 "evidence": "llama_cpp GGUF + LM Studio local server, no cloud SDK",1132 "source": "app_file"1133 }1134 ],1135 "signal_profile": "both",1136 "audit": "KEPT Llama Champion (literal import + *Q4_K_M.gguf). KEPT Backyard AI, trimmed 0.85 (group not named person). KEPT Tiny Titan: 3 models all <=4B (3.35B max). KEPT Off the Grid 0.75: llama.cpp/GGUF + local LM Studio; no cloud SDK (Cohere is open-weight). DROPPED Off-Brand: Catpuccin theme is a docstring comment, not real CSS code."1137 },1138 {1139 "slug": "wonderland",1140 "matches": [1141 {1142 "quest": "Nemotron",1143 "confidence": 0.97,1144 "evidence": "MODEL_ID = \"nvidia/NVIDIA-Nemotron-3-Nano-4B-BF16\"",1145 "source": "app_file"1146 },1147 {1148 "quest": "Tiny Titan",1149 "confidence": 0.94,1150 "evidence": "Nemotron 3 Nano 4B (~4B params)",1151 "source": "app_file"1152 },1153 {1154 "quest": "Off the Grid",1155 "confidence": 0.9,1156 "evidence": "running fully locally; AutoModelForCausalLM, no cloud API",1157 "source": "app_file"1158 },1159 {1160 "quest": "Thousand Token Wood",1161 "confidence": 0.93,1162 "evidence": "A tiny text adventure; the forgetting is the game",1163 "source": "app_file"1164 }1165 ],1166 "signal_profile": "app_only",1167 "audit": "KEPT 4. Nemotron: exact nvidia/...Nemotron MODEL_ID. Tiny Titan: active model is real 4B Nano (ignored aspirational 30B comment). Off the Grid: local transformers, no cloud API. TTW: text-adventure framing. Did NOT add Llama Champion (GGUF only an aspirational comment; real load is BF16 transformers, no llama_cpp), Best Agent ('agentic' marketing), Field Notes (README boilerplate). app_only."1168 },1169 {1170 "slug": "awaaz",1171 "matches": [1172 {1173 "quest": "Off the Grid",1174 "confidence": 0.78,1175 "evidence": "transformers AutoModelForCausalLM load, no openai/anthropic client",1176 "source": "app_file"1177 },1178 {1179 "quest": "Backyard AI",1180 "confidence": 0.85,1181 "evidence": "dub into the natural Hindi mix your family actually uses",1182 "source": "readme"1183 }1184 ],1185 "signal_profile": "both",1186 "audit": "KEPT Off the Grid: app_file loads transformers AutoModelForCausalLM, no cloud LLM client visible (Edge TTS is free TTS); conf lowered since APP_FILE truncated. KEPT Backyard AI: README frames concrete household problem; re-anchored off bare track-tag line. DROPPED Field Notes: short pipeline list is a blurb, not a substantial build write-up. No Tiny Titan (Qwen 7B, over 4B)."1187 },1188 {1189 "slug": "karim-lab",1190 "matches": [],1191 "signal_profile": "none",1192 "audit": "Dropped Backyard AI: quote is real but persona is \"a lawyer\" (generic profession, not a named person/household task), so the bar fails; in doubt, drop. No other quest holds: app is deterministic regex with NO model (\"does not call a hosted LLM yet\"), so Off the Grid fails. llama.cpp/Modal/Field Notes only PLANNED. declared_models empty (no Tiny Titan). Plain gradio, no theming/agent loop."1193 },1194 {1195 "slug": "trace-field-notes",1196 "matches": [1197 {1198 "quest": "OpenBMB",1199 "confidence": 0.95,1200 "evidence": "openbmb/MiniCPM5-1B pass by default",