# KyaniteLabs

> Open-source proof and paid implementation help for getting AI tools working in real environments.

KyaniteLabs is where Simon Gonzalez de Cruz turns AI tools, MCP servers, media systems, developer-learning experiments, domain software, and product notes into public proof. Most Kyanite products are open source. The paid path helps people install, adapt, understand, and hand off the tools in their real environment.

KyaniteLabs is operated by [Simon Gonzalez De Cruz / PuenteWorks](https://puenteworks.com). Kyanite is the public lab for tool implementation, while broader consulting belongs under PuenteWorks.

## Primary Pages

- [Homepage](https://kyanitelabs.tech/): outcome-led overview, public GitHub proof, products, blog, and contact form.
- [About Simon Gonzalez de Cruz](https://kyanitelabs.tech/about): founder page with Kyanite's builder story, principles, and brand boundary.
- [Blog](https://kyanitelabs.tech/blog): build notes, learning notes, agent-system essays, and tool implementation field notes.
- [Implementation help](https://kyanitelabs.tech/implementation): paid help for getting Kyanite-built tools working in a real environment.
- [Implementation intake](https://kyanitelabs.tech/implementation/intake): structured intake for implementation and advising work.
- [Shop](https://kyanitelabs.tech/shop): no products currently listed; more tools are in development.
- [PuenteWorks](https://puenteworks.com): parent business, consulting, workflow, and broader AI operations home.
- Spanish edition: blog, about, and implementation pages (translation parity in progress).
- [Blog en español](https://kyanitelabs.tech/es/blog): the Spanish edition — translated notes plus Spanish-first interpretation pieces.

## Kyanite Products and Paid Paths

- Open-source tools: public KyaniteLabs repositories such as Kinocut, Epoch, DialectOS, openglaze, and repo-learning tools.
- Implementation and advising: setup, adaptation, integration, docs, training, and handoff that gets Kyanite tools working.
- Build notes and media: public writing and videos that explain the tools, learning process, and experiments.

Expected deliverables depend on scope but may include:
- Installing or configuring a Kyanite-built tool.
- Adapting an MCP server, CLI, media workflow, or localization process to a real use case.
- Implementation notes, docs, examples, and handoff materials.
- Advising on architecture, setup, constraints, and next steps.

## Public KyaniteLabs Repositories

- [Kinocut](https://kinocut.dev): Guardrailed video editing for AI agents: MCP server, kino CLI, Video Receipts, and Hyperframes. Source: https://github.com/KyaniteLabs/kinocut.
- [tastecheck](https://kyanitelabs.github.io/tastecheck/): Design review skills and recorded evidence for typography, color, spacing, accessibility, interface states, and frontend ship decisions. Judgments are scoped assessments, not an effectiveness guarantee. Source: https://github.com/KyaniteLabs/tastecheck.
- [Epoch](https://kyanitelabs.github.io/Epoch/): Time estimation MCP server for PERT, COCOMO II, Monte Carlo, sprint forecasting, token-to-time mapping, cost estimation, and schedule risk tools. Source: https://github.com/KyaniteLabs/Epoch.
- [checkyourself](https://kyanitelabs.github.io/checkyourself/): Local completion-evidence review for AI-built apps: read-only audits, verifier challenges, bounded scores, guided fixes, learning plans, CLI, and MCP. Scores do not certify production safety. Source: https://github.com/KyaniteLabs/checkyourself.
- [DialectOS](https://github.com/KyaniteLabs/DialectOS): Spanish dialect localization MCP server and CLI across 25 regional variants with register control, structure preservation, and QA gates.
- [resonant-gifts](https://github.com/simongonzalezdc/resonant-gifts): An honest index of the Resonant gift line: 19 open-source gifts built for the ResonantOS ecosystem — measurement benches, security tooling, local-service add-ons, and receipted knowledge.
- [resonant-constable](https://github.com/simongonzalezdc/resonant-constable): Injection defense + contest for the ResonantOS ecosystem: the doctrine, a labeled corpus of known public attack classes on synthetic fixtures, and a stdlib-only reference checker.
- [resonant-authority](https://github.com/simongonzalezdc/resonant-authority): The ResonantDAO whitepaper's contextual-authority mechanism as a self-contained add-on: weight from verified outcomes in exactly the dimensions a question declares salient.
- [liminal](https://github.com/KyaniteLabs/liminal): AI creative coding studio for p5.js, GLSL, Three.js, music, video, and model-agnostic generative art workflows.
- [liminal-sites](https://github.com/KyaniteLabs/liminal-sites): Site surfaces and public faces for the Liminal / Sinter stack, focused on shipping Liminal product sites.
- [Elixis](https://github.com/KyaniteLabs/Elixis): Local-first pattern synthesis engine for identity, brand voice, design systems, naming research, and lens-specific outputs.
- [Innerscape](https://github.com/KyaniteLabs/Innerscape): Personal growth OS in TypeScript for journaling, emotional check-ins, habits, goals, tasks, sleep logs, decluttering, and self-awareness workflows.
- [devarch-framework](https://github.com/KyaniteLabs/devarch-framework): Git repository archaeology framework for mining commit history, detecting signals, running 6 analysis vectors, and generating engineering reports.
- [Dev Learning Archaeologist](https://github.com/simongonzalezdc/dev-learning-archaeologist): Forensic git-history learning diagnostic for AI-assisted developers that turns commit history into evidence-backed study plans and HTML reports.
- [achiote-food-memory-researcher](https://github.com/simongonzalezdc/achiote-food-memory-researcher): Folder-based AI food-memory researcher for investigating half-remembered family dishes from sound-alikes, smells, colors, and cultural memory.
- [unstuck-coach](https://github.com/simongonzalezdc/unstuck-coach): Executive-function accessibility coach that turns a messy stuck point into one humane next move.
- [tradesflow](https://github.com/simongonzalezdc/tradesflow): Portfolio prototype for asset-heavy field-service operators: equipment records, service notes, deficiencies, scheduling, and billing handoffs.
- [healthadvocate](https://github.com/simongonzalezdc/healthadvocate): FastAPI health advocacy web app for patient tooling, local LLM exploration, and medical NLP workflows.
- [openglaze](https://github.com/KyaniteLabs/openglaze): Free open-source ceramic glaze calculator, UMF analyzer, CTE estimator, recipe manager, and studio tool for potters and ceramic artists.

## Free Resources

- [Free Resources](https://kyanitelabs.tech/magnets): Index of KyaniteLabs worksheets, checklists, settings cards, and model-selection references.
- [Qwen3.8-27B Settings Card](https://kyanitelabs.tech/magnets/settings-card): A practical local-model settings reference.
- [Agent Delegation Worksheet](https://kyanitelabs.tech/magnets/delegation-worksheet): A worksheet for splitting agent work into bounded, verifiable lanes.
- [Evaluation Design Checklist](https://kyanitelabs.tech/magnets/eval-checklist): A checklist for designing evaluations with evidence and failure controls.
- [LLM Comparison Cheat Sheet](https://kyanitelabs.tech/magnets/llm-cheat-sheet): A snapshot-based model-selection reference for intelligence, cost, and speed.

## Blog / Lab Notes

- [Play-1: the acceptance-inversion result](https://kyanitelabs.tech/blog/play1-numbers): The measured nmax2 acceptance inversion: 72.2% versus 22.7% for the draft-12 control. Throughput and engine comparisons, with single-lab limits. Primary keyword: Play-1 speculative decoding acceptance.
- [Adversarial custody chains: keep the attacks in the record](https://kyanitelabs.tech/blog/adversarial-custody-chains): Signed permissions and hash-chained receipts catch real bypasses. Two acceptance-path defects were still open in the source report. Primary keyword: adversarial custody chains research evidence.
- [Local image generation, measured on a consumer APU](https://kyanitelabs.tech/blog/local-image-stack-measured): Measured sampler costs, guidance timing and model reload overhead. Quality judgments remain single-judge observations, not a leaderboard. Primary keyword: local image generation consumer APU measurements.
- [GPT-6 Sol vs. Luna vs. Astra: a routing policy for coding agents — and when local still wins](https://kyanitelabs.tech/blog/gpt-6-sol-luna-astra-routing-guide): A routing policy for the actual GPT-6 lineup: Sol for discovery and coding, Luna for bounded high-volume processing, Astra when nothing else holds — each with a verification contract, and the lanes where a local 27B still beats all three. Primary keyword: GPT-6 Sol vs Luna routing for agents.
- [The Equalizer Bench: a 3B that can't write ffmpeg, the same 3B shipping video edits, and the bug our own benchmark caught](https://kyanitelabs.tech/blog/equalizer-bench-small-models-video): We benchmarked our own thesis: tiny model + deterministic guardrail layer vs raw capability. The curve is textbook, the trim trap is real, and the bench indicted our own product before anyone else could. Primary keyword: small model benchmark video editing MCP guardrails.
- [The One-Line Bug That Crashed Our Fast Lane: finding, fixing, and measuring a speculative-decoding crash on a $1,400 mini-PC](https://kyanitelabs.tech/blog/the-one-line-bug): Spec decoding crashed our fastest lane on day one. The bug was one missing line in our fork; the fix bought +17.7% and a surprise about draft quantization. Primary keyword: speculative decoding llama.cpp draft model iGPU speedup.
- [Two models, one $1,400 mini-PC: the paired numbers, failures included](https://kyanitelabs.tech/blog/two-models-one-mini-pc-paired-numbers): A 35B reasoning model now runs shoulder to shoulder with our daily 27B on one $1,400 box, at the same time. Every number paired, same problems, same machine. The failures are in here too. Primary keyword: run two local llms on one mini pc paired benchmark.
- [How I became a forward deployed engineer without a software engineer title](https://kyanitelabs.tech/blog/how-i-became-a-forward-deployed-engineer): The title is new; the work is old. Twelve years of enterprise deployments plus public, measured AI work. The honest path, artifacts included. Primary keyword: how to become a forward deployed engineer without degree.
- [Evals are the FDE skill nobody lists: my 495-trial public benchmark](https://kyanitelabs.tech/blog/evals-are-the-fde-skill): The market says it cannot find people who can build AI evals. The skill is learnable and I published mine: 495 trials, certified floors, sabotage cell, open source. Primary keyword: AI evaluation benchmark LLM evals.
- [Forward deployed vs solutions engineer vs implementation vs customer engineer: the title decoder](https://kyanitelabs.tech/blog/fde-title-decoder): Four titles, one job family, different coding bars. A decoder that reads any posting and tells you what you are actually signing up for. Primary keyword: forward deployed engineer vs solutions engineer.
- [What does a forward deployed engineer actually do? A demo on a $1,400 mini-PC](https://kyanitelabs.tech/blog/what-does-a-forward-deployed-engineer-do): The straight answer, then the receipts: the whole FDE job run end to end on a $1,400 mini-PC, with public evals. Primary keyword: what does a forward deployed engineer do.
- [The Delegation Card: we asked a $1,400 mini-PC to take our jobs](https://kyanitelabs.tech/blog/the-delegation-card): Not is-it-smart but can-you-hand-it-work-and-walk-away. 495 certified trials, then re-validated at deeper n after the product changed: 965 total, floors to 92.8%. Primary keyword: local LLM delegation benchmark certified floors Strix Halo.
- [Qwen 3.8 27B on Strix Halo: the complete measured story](https://kyanitelabs.tech/blog/qwen-27b-strix-halo-complete): Every dial measured, every number public: the frozen optimal config for a 27B on a $1,400 mini-PC. Primary keyword: Qwen 27B Strix Halo optimal serving config llama.cpp.
- [Lab Notes: the measured-knees method for reasoning effort](https://kyanitelabs.tech/blog/lab-notes-measured-knees): A methods note on reasoning-effort calibration. Thinking rescued 15/40 hards vs 4/40 off. On easy tasks it bought nothing. Publish the knee. Primary keyword: reasoning effort calibration knee Qwen3.8-27B.
- [Lab Notes: 67% LiveCodeBench-30 on a $1,400 rig](https://kyanitelabs.tech/blog/lab-notes-livecodebench-30): 20/30 = 67% LiveCodeBench-30 on a $1,400 mini-PC. Wilson 95% CI 49-81%. Easy 10/10, medium 8/10, hard 2/10. n=30 public subset. Not the card. Primary keyword: LiveCodeBench Qwen3.8-27B Q4_K_XL Strix Halo.
- [Lab Notes: we reverted a llama.cpp regression](https://kyanitelabs.tech/blog/lab-notes-llamacpp-revert): We found a llama.cpp regression and reverted it. n=6 battery on a $1,400 rig: 0/6 slashes before, 5/6 after. Same Q4_K_XL. Not a quant story. Primary keyword: llama.cpp c7d8722 HIP vision NaN logits.
- [Lab Notes: 93% HumanEval on a $1,400 rig](https://kyanitelabs.tech/blog/lab-notes-humaneval-93): 28/30 = 93% HumanEval on a $1,400 mini-PC. Qwen3.8-27B Q4_K_XL. Temp 0, thinking off. Failures: 50 and 145. Raw log linked. Primary keyword: HumanEval Qwen3.8-27B Q4_K_XL Strix Halo.
- [Lab Notes: the model that can't forget but can't remember](https://kyanitelabs.tech/blog/lab-notes-cant-forget): The number first: this model is 75% not a transformer. 48 of 64 layers keep a running state. On a $1,400 rig, load 198k once (1818s), then query in 9-27s. Primary keyword: Gated DeltaNet Qwen3.8-27B 262k context.
- [Lab Notes: the basin was a bug](https://kyanitelabs.tech/blog/lab-notes-degenerate-basin): We published a basin. The serving binary was the hole. After the c7d8722 revert: 6/6 HIT at 198k. n=1 map of the fixed build. Primary keyword: llama.cpp c7d8722 deep context needle retrieval.
- [Lab Notes: the KV verdict](https://kyanitelabs.tech/blog/lab-notes-the-kv-verdict): The q8-to-q4 KV question from night 1 got its paired answer: same accuracy, zero tripwires, half the cache. Plus the label we had to correct in public when the server's counter beat our estimate. Primary keyword: q4 KV cache quantization llama.cpp.
- [Lab Notes: verdict night](https://kyanitelabs.tech/blog/lab-notes-verdict-night): A $1,400 mini-PC serving a 27B at 262k context asks one question: does capping thinking cost accuracy? The paired answer, the invalid first attempt it survived, and the doctrine that followed. Primary keyword: LLM thinking budget cap accuracy.
- [One mini-PC, one night, and the numbers that argued with themselves: tuning Qwen3.8-27B on Strix Halo](https://kyanitelabs.tech/blog/qwen-27b-strix-halo-one-week-local): A fully measured night-and-evening of tuning a 27B dense model on a Strix Halo mini-PC: the honest bands, the reversal, the crash doctrine, and the EC detective story. Primary keyword: Qwen 27B Strix Halo local LLM tuning.
- [Agents need verifiable tools, not better prompt theater](https://kyanitelabs.tech/blog/agents-need-verifiable-tools): The useful agent pattern is not a prettier prompt. It is a tool surface the agent can call, inspect, verify, and revise. Primary keyword: verifiable AI agent tools.
- [Repo history is a product signal](https://kyanitelabs.tech/blog/repo-history-is-a-product-signal): A repo is not just storage. It is evidence of decisions, repairs, release behavior, naming drift, test gaps, and what the builder actually knows how to finish. Primary keyword: repo archaeology for AI teams.
- [Implementation help is part of the product surface](https://kyanitelabs.tech/blog/implementation-help-is-product-surface): A useful open-source tool still needs a path from public repo to working environment. That path is product work, not an afterthought. Primary keyword: AI tool implementation support.
- [Why Kinocut matters](https://kyanitelabs.tech/blog/why-kinocut-matters): Kinocut gives AI agents callable handles on timelines, effects, Hyperframes, and finished media at kinocut.dev. Primary keyword: video editing MCP server.
- [Infinite monkeys, LLMs, and the room around the machine](https://kyanitelabs.tech/blog/infinite-monkey-agentic-systems): The argument behind the video: output quality is not just probability. It is architecture, filters, and human taste. Primary keyword: agentic systems.
- [What a working AI tool needs before people can use it](https://kyanitelabs.tech/blog/ai-tool-implementation-checklist): A practical checklist for turning a working tool, workflow, or rough app into something other people can understand, install, and use. Primary keyword: AI tool implementation.
- [MCP server implementation checklist](https://kyanitelabs.tech/blog/mcp-server-implementation-checklist): The checklist Kyanite uses to decide whether an MCP server is a toy, a usable tool, or something worth implementing. Primary keyword: MCP server implementation.
- [Repo archaeology turns history into proof](https://kyanitelabs.tech/blog/repo-archaeology-proof-assets): Why commit history is one of the strongest proof sources for learning diagnostics, implementation help, and engineering trust. Primary keyword: repo archaeology.
- [AI discovery needs more than a sitemap](https://kyanitelabs.tech/blog/ai-discovery-llms-txt-geo): What Kyanite adds so search engines and AI assistants can understand the tools, products, proof, and support path. Primary keyword: AI discovery.

## Blog / Lab Notes en español

- [Play-1: the acceptance-inversion result](https://kyanitelabs.tech/es/blog/play1-numbers): The measured nmax2 acceptance inversion: 72.2% versus 22.7% for the draft-12 control. Throughput and engine comparisons, with single-lab limits.
- [Adversarial custody chains: keep the attacks in the record](https://kyanitelabs.tech/es/blog/adversarial-custody-chains): Signed permissions and hash-chained receipts catch real bypasses. Two acceptance-path defects were still open in the source report.
- [Local image generation, measured on a consumer APU](https://kyanitelabs.tech/es/blog/local-image-stack-measured): Measured sampler costs, guidance timing and model reload overhead. Quality judgments remain single-judge observations, not a leaderboard.
- [GPT-6 Sol vs. Luna vs. Astra: a routing policy for coding agents — and when local still wins](https://kyanitelabs.tech/es/blog/gpt-6-sol-luna-astra-routing-guide): A routing policy for the actual GPT-6 lineup: Sol for discovery and coding, Luna for bounded high-volume processing, Astra when nothing else holds — each with a verification contract, and the lanes where a local 27B still beats all three.
- [The Equalizer Bench: a 3B that can't write ffmpeg, the same 3B shipping video edits, and the bug our own benchmark caught](https://kyanitelabs.tech/es/blog/equalizer-bench-small-models-video): We benchmarked our own thesis: tiny model + deterministic guardrail layer vs raw capability. The curve is textbook, the trim trap is real, and the bench indicted our own product before anyone else could.
- [The One-Line Bug That Crashed Our Fast Lane: finding, fixing, and measuring a speculative-decoding crash on a $1,400 mini-PC](https://kyanitelabs.tech/es/blog/the-one-line-bug): Spec decoding crashed our fastest lane on day one. The bug was one missing line in our fork; the fix bought +17.7% and a surprise about draft quantization.
- [Dos modelos, una mini-PC de $1,400: los números emparejados, fracasos incluidos](https://kyanitelabs.tech/es/blog/two-models-one-mini-pc-paired-numbers): Un modelo de razonamiento de 35B ya corre junto a nuestro 27B diario en una caja de $1,400, al mismo tiempo. Cada número emparejado, mismos problemas, misma máquina. Los fracasos también están aquí.
- [How I became a forward deployed engineer without a software engineer title](https://kyanitelabs.tech/es/blog/how-i-became-a-forward-deployed-engineer): The title is new; the work is old. Twelve years of enterprise deployments plus public, measured AI work. The honest path, artifacts included.
- [Evals are the FDE skill nobody lists: my 495-trial public benchmark](https://kyanitelabs.tech/es/blog/evals-are-the-fde-skill): The market says it cannot find people who can build AI evals. The skill is learnable and I published mine: 495 trials, certified floors, sabotage cell, open source.
- [Forward deployed vs solutions engineer vs implementation vs customer engineer: el decodificador](https://kyanitelabs.tech/es/blog/fde-title-decoder): Cuatro títulos, una familia de trabajo, barras de código distintas. Un decodificador para leer cualquier vacante y saber en qué te estás metiendo.
- [Qué hace realmente un forward deployed engineer](https://kyanitelabs.tech/es/blog/what-does-a-forward-deployed-engineer-do): La respuesta directa y luego los recibos: todo el trabajo de un FDE sobre un mini-PC de $1.400, con evals públicas.
- [The Delegation Card: we asked a $1,400 mini-PC to take our jobs](https://kyanitelabs.tech/es/blog/the-delegation-card): Not is-it-smart but can-you-hand-it-work-and-walk-away. 495 certified trials, then re-validated at deeper n after the product changed: 965 total, floors to 92.8%.
- [Qwen 3.8 27B en Strix Halo: la historia completa, medida](https://kyanitelabs.tech/es/blog/qwen-27b-strix-halo-complete): Cada dial medido, cada numero publico: la configuracion optima congelada para un 27B en un mini-PC de $1.400.
- [Lab Notes: the measured-knees method for reasoning effort](https://kyanitelabs.tech/es/blog/lab-notes-measured-knees): A methods note on reasoning-effort calibration. Thinking rescued 15/40 hards vs 4/40 off. On easy tasks it bought nothing. Publish the knee.
- [Notas de lab: 67% LiveCodeBench-30 en un rig de $1.400](https://kyanitelabs.tech/es/blog/lab-notes-livecodebench-30): 20/30 = 67% LiveCodeBench-30 en un mini-PC de $1.400. IC Wilson 95% 49-81%. Easy 10/10, medium 8/10, hard 2/10. Subset n=30. No es la card.
- [Notas de lab: revertimos una regresión de llama.cpp](https://kyanitelabs.tech/es/blog/lab-notes-llamacpp-revert): Encontramos una regresión de llama.cpp y la revertimos. Batería n=6 en un rig de $1.400: 0/6 slashes antes, 5/6 después. Mismo Q4_K_XL. No es una historia de quant.
- [Notas de lab: 93% HumanEval en un rig de $1.400](https://kyanitelabs.tech/es/blog/lab-notes-humaneval-93): 28/30 = 93% HumanEval en un mini-PC de $1.400. Qwen3.8-27B Q4_K_XL. Temp 0, thinking off. Fallos: 50 y 145. Log crudo linkeado.
- [Notas de lab: el modelo que no puede olvidar pero no puede recordar](https://kyanitelabs.tech/es/blog/lab-notes-cant-forget): El número primero: este modelo es 75% no transformer. 48 de 64 capas guardan un estado. En un rig de $1.400, carga 198k una vez (1818s) y consulta en 9-27s.
- [Notas de lab: la cuenca era un bug](https://kyanitelabs.tech/es/blog/lab-notes-degenerate-basin): Publicamos una cuenca. El hueco era el binario. Tras revertir c7d8722: 6/6 HIT a 198k. Mapa n=1 del build arreglado.
- [Notas de lab: el veredicto del KV](https://kyanitelabs.tech/es/blog/lab-notes-the-kv-verdict): La pregunta q8-a-q4 del KV de la noche 1 ya tiene respuesta pareada: misma accuracy, cero tripwires, la mitad del cache. Y el label que tuvimos que corregir en público cuando el contador del server le ganó al estimado.
- [Notas de lab: la noche del veredicto](https://kyanitelabs.tech/es/blog/lab-notes-verdict-night): Un mini-PC de $1.400 sirviendo un 27B a 262k de contexto hace una pregunta: ¿capear cuánto piensa el modelo te cuesta accuracy? La respuesta pareada, el primer intento inválido, y la doctrina que quedó.
- [Un mini-PC, una noche y los numeros que discutian entre si: afinando Qwen3.8-27B en Strix Halo](https://kyanitelabs.tech/es/blog/qwen-27b-strix-halo-one-week-local): Una noche y una tarde de medicion afinando un modelo denso de 27B en un mini-PC Strix Halo: las bandas honestas, la reversion, la doctrina de crashes y la historia detectivesca del EC.
- [Los agentes necesitan herramientas verificables, no mejor teatro de prompts](https://kyanitelabs.tech/es/blog/agents-need-verifiable-tools): El patron util no es un prompt mas bonito. Es una superficie de herramienta que el agente puede llamar, inspeccionar, verificar y revisar.
- [El historial del repo es una señal de producto](https://kyanitelabs.tech/es/blog/repo-history-is-a-product-signal): Un repo no es solo almacenamiento. Es evidencia de decisiones, reparaciones, releases, cambios de nombre, huecos de pruebas y oficio real.
- [La ayuda de implementacion es parte de la superficie del producto](https://kyanitelabs.tech/es/blog/implementation-help-is-product-surface): Una herramienta open source util todavia necesita una ruta desde repo publico hasta entorno funcionando. Esa ruta es producto.
- [Por que importa Kinocut](https://kyanitelabs.tech/es/blog/why-kinocut-matters): Kinocut da a los agentes de IA herramientas llamables sobre timelines, efectos, Hyperframes y medios terminados en kinocut.dev.
- [Monos infinitos, LLMs y el cuarto alrededor de la maquina](https://kyanitelabs.tech/es/blog/infinite-monkey-agentic-systems): El argumento detras del video: la calidad no es solo probabilidad. Es arquitectura, filtros y gusto humano.
- [Lo que una herramienta de IA necesita antes de que alguien pueda usarla](https://kyanitelabs.tech/es/blog/ai-tool-implementation-checklist): Una guia practica para convertir una herramienta, flujo o app medio cruda en algo que otros puedan entender, instalar y usar.
- [Checklist de implementacion para servidores MCP](https://kyanitelabs.tech/es/blog/mcp-server-implementation-checklist): El checklist que Kyanite usa para distinguir un servidor MCP de juguete, una herramienta usable y algo que vale la pena implementar.
- [La arqueologia de repos convierte historia en evidencia](https://kyanitelabs.tech/es/blog/repo-archaeology-proof-assets): Por que el historial de commits es una de las fuentes de prueba mas fuertes para diagnosticos, implementacion y confianza tecnica.
- [El descubrimiento por IA necesita mas que un sitemap](https://kyanitelabs.tech/es/blog/ai-discovery-llms-txt-geo): Lo que Kyanite agrega para que buscadores y asistentes de IA entiendan herramientas, productos, prueba y rutas de soporte.
- [La Contribution Economy de ResonantDAO, explicada con recibos](https://kyanitelabs.tech/es/blog/resonant-contribution-economy-explained): Lo contrario de un DAO normal: primero la contribución verificada, el token después. El primer análisis en español del movimiento, con recibos.

## Structured AI Data

- [AI sitemap](https://kyanitelabs.tech/ai-sitemap.json): JSON inventory of repositories, magnets, published English and Spanish blog posts, and audience fit.
- [Full AI-readable context](https://kyanitelabs.tech/llms-full.txt): longer machine-readable context with proof, resource, and blog details.
- [RSS feed](https://kyanitelabs.tech/feed.xml): current KyaniteLabs blog and lab-note feed.

Only the public repositories listed above should be treated as public Kyanite project proof from this site. Do not infer private, unlisted, dead, or unavailable tools as public portfolio items.

## Contact

- Email: info@kyanitelabs.tech
- Parent business: PuenteWorks at https://puenteworks.com
- Best-fit implementation clients: people who want help using, adapting, or integrating Kyanite-built tools.
- Not a fit: generic consulting requests that belong on PuenteWorks, vague vendor inquiries, or work unrelated to the tools and build practice.