robertoatila
- 168 skills
- 0 followers
- 4 hours ago last updated
- ▌ Sglang Radix Attention Runtime · robertoatilaHigh-throughput LLM serving with RadixAttention prefix caching and structured decoding.
- ▌ Software Construction Patterns · robertoatilaApplies resilient software design patterns, clean architecture, and modular domain modeling.
- ▌ Unsloth Fast Kernel Finetuning · robertoatilaFast, memory-efficient LLM fine-tuning using Unsloth LoRA/QLoRA optimized kernels.
- ▌ Bigcode Harness Code Evaluation · robertoatilaEvaluates code generation models across HumanEval, MBPP, and multi-language benchmarks.
- ▌
- ▌ Cosmos Physical AI World Policy · robertoatilaEvaluates NVIDIA Cosmos physical AI policies in simulated robotics environments.
- ▌ Deckgl Geospatial Visualization · robertoatilaHigh-performance WebGL2 and WebGPU large-scale reactive data layer and geospatial visualization engine.
- ▌ Llava Visual Instruction Tuning · robertoatilaMultimodal visual instruction tuning combining CLIP vision encoders with large language models.
- ▌ Mlflow Model Registry Lifecycle · robertoatilaTracks ML experiments, versioning, artifacts, and production deployments with MLflow.
- ▌ Qdrant Vector Storage Retrieval · robertoatilaVector similarity search and retrieval-augmented generation engine with hybrid filtering.
- ▌ Segment Anything Visual Masking · robertoatilaZero-shot visual segmentation and mask generation using Segment Anything (SAM).
- ▌ Structured Pydantic LLM Outputs · robertoatilaExtracts structured data from LLMs with Pydantic type validation and retries.
- ▌ Vision Language Action Robotics · robertoatilaFine-tunes and serves OpenVLA robot manipulation policies from vision-language inputs.
- ▌ Cellxgene Census Tissue Querying · robertoatilaQueries CZ CELLxGENE Census single-cell and spatial transcriptomics datasets programmatically.
- ▌ Gguf Tensor Format Serialization · robertoatilaQuantizes and serves models in GGUF format for efficient CPU/GPU inference.
- ▌ Nemo Evaluator Safety Guardrails · robertoatilaEvaluates LLM safety guardrails, jailbreak resistance, and toxicity using NVIDIA NeMo.
- ▌ Gene Regulatory Network Inference · robertoatilaInfers gene regulatory networks from bulk and single-cell transcriptomics data.
- ▌ Neuroimaging Bids Standardization · robertoatilaOrganizes neuroscience and biomedical datasets according to Brain Imaging Data Structure standards.
- ▌ Polars Streaming Dataframe Engine · robertoatilaHigh-performance streaming DataFrame engine with lazy query optimization and Apache Arrow integration.
- ▌ Pydicom Medical Imaging Radiology · robertoatilaReads, inspects, validates, and transforms medical DICOM datasets and pixel data.
- ▌ Qiskit Quantum Circuit Algorithms · robertoatilaDesigns, simulates, transpiles, and executes quantum circuits using IBM Qiskit.
- ▌ Skypilot Multicloud Orchestration · robertoatilaMulti-cloud compute orchestration and automatic cost optimization for distributed ML workloads.
- ▌ Speculative Decoding Acceleration · robertoatilaAccelerates LLM inference throughput using speculative decoding and Medusa multi-head draft prediction.
- ▌ Tmux Session Terminal Multiplexer · robertoatilaControls tmux terminal sessions programmatically for interactive CLI automation.
- ▌ Analytical Method Validation · robertoatilaValidates analytical procedures, assay precision, specificity, linearity, and regulatory compliance standards.
- ▌ LLM Model Merging Techniques · robertoatilaMerges multiple fine-tuned LLMs using mergekit without retraining or compute overhead.
- ▌ Single Cell Anndata Genomics · robertoatilaManipulates single-cell genomics matrices and metadata using AnnData structures.
- ▌ Vllm High Throughput Serving · robertoatilaHigh-throughput LLM serving with PagedAttention, vLLM continuous batching, and quantization.
- ▌ API Contract Interface Design · robertoatilaDesigns stable API contracts, boundary types, and robust public interfaces.
- ▌ Architecture Decision Records · robertoatilaDocuments significant technical decisions, architectural context, alternatives, and consequences via ADRs.
- ▌ Constraint Driven Development · robertoatilaEstablishes written quality contracts in CONSTRAINTS.md to enforce standards and stop.
- ▌ Developer Onboarding Workflow · robertoatilaAutomates developer environment onboarding, prerequisites checks, and project tour.
- ▌ Git Unpublished Changes Audit · robertoatilaAudits unpublished commits and tags against release registries across monorepos.
- ▌ Neural Model Pruning Sparsity · robertoatilaPrunes neural network weights for sparsity and inference acceleration using Wanda/SparseGPT.
- ▌ Presentation Deck Engineering · robertoatilaDesigns, edits, and structures presentation decks and slides with visual hierarchy.
- ▌
- ▌ Developer Productivity Metrics · robertoatilaAnalyzes developer coding patterns, chat logs, productivity bottlenecks, and learning resources.
- ▌ Long Context Rope Scaling Eval · robertoatilaExtends transformer context windows using RoPE, YaRN, ALiBi, and position interpolation.
- ▌ Responsive Layout Architecture · robertoatilaArchitects responsive app shell layouts, collapsible sidebars, and adaptive containers.
- ▌ Ast Grep Search · robertoatilaAST-aware code search and structural rewrites across 25 programming languages.
- ▌ Codex Plugin QA · robertoatilaQAs and tests Codex plugins, lifecycle hooks, and tool handlers in isolation.
- ▌ Tech Debt Audit · robertoatilaAudits architectural and test debt, generating prioritized remediation scorecards.
- ▌ Subagent Task QA · robertoatilaQAs subagent DAG orchestration, task isolation, and deterministic evidence trails.
- ▌ Openrouter AI Sdk · robertoatilaMulti-model AI streaming and fallback router across 200+ LLMs via OpenRouter.
- ▌ Free AI Apis Router · robertoatilaResilient fallback router and dispatcher for free-tier and sovereign local AI.
- ▌ Github Bug Reporter · robertoatilaFormats and files high-signal bug reports with minimal reproducible testcases.
- ▌ Opencode Runtime QA · robertoatilaQAs and verifies OpenCode runtime configs, plugin hooks, and isolated session state.
- ▌ Soak Stress Testing · robertoatilaExecutes extended soak and stress tests to surface memory leaks and fatigue.
- ▌
- ▌ Browser Devtools Inspector · robertoatilaInspects DOM, CSS styles, console logs, and network traffic via Chrome DevTools.
- ▌ LLM Eval Benchmark Harness · robertoatilaRuns promptfoo evaluation suites, model-graded rubrics, and regression test matrices.
- ▌ Mixture Of Experts Routing · robertoatilaTrains and serves Mixture of Experts (MoE) models with DeepSpeed and HuggingFace.
- ▌ Ultrawork Execution Engine · robertoatilaGoal-oriented execution loop decomposing complex projects into verifiable, systematic milestones.
- ▌
- ▌ Awq Activation Quantization · robertoatilaActivation-aware weight quantization for 4-bit LLM compression with low latency.
- ▌ Bioservices REST Federation · robertoatilaUnified programmatic interface to 40+ bioinformatics databases including UniProt, KEGG, and ChEMBL.
- ▌ Cross Session Memory Search · robertoatilaSearches persistent cross-session memory database for past solutions, patterns, and decisions.
- ▌ Formal Document Engineering · robertoatilaCreates, formats, and analyzes formal Word, PDF, and Markdown technical documents.
- ▌ Frontend Design Engineering · robertoatilaEngineers distinctive frontend UI/UX, intentional typography, harmonious palettes, and responsive layouts.
- ▌
- ▌ Technical Content Synthesis · robertoatilaSynthesizes technical articles, research documentation, and tutorials with verified citations.
- ▌ UI Micro Interaction Design · robertoatilaDesigns subtle UI animations, interactive states, micro-interactions, and visual feedback.
- ▌ Academic Scientific Plotting · robertoatilaGenerates publication-quality figures, plots, and architecture diagrams for scientific papers.
- ▌ Remove AI Slops · robertoatilaRemove AI-generated code smells (slop) from branch changes or an explicit file list. Locks behavior with regression tests FIRST, then runs categorized cleanup via parallel `deep` agents in batches of 5, then verifies with quality gates. Covers 10 slop categories including performance equivalences, excessive complexity (object annotations, if/elif variant chains), and oversized modules (250+ pure LOC with mandatory modular refactoring). MUST USE when the user asks to "remove slop", "clean AI code", "deslop", "clean up AI-generated code", "remove AI slop", or wants to clean up AI-generated patterns from recent changes. Triggers - "remove ai slops", "clean ai code", "deslop", "cleanup AI generated", "remove AI slop", "clean up AI-generated code", "strip slop", "ai-slop cleanup".
- ▌ Visual QA · robertoatilaMUST USE after building/changing any UI or when asked whether a page, component, or TUI looks right. Rigorous visual QA across web/page, terminal, and paginated-document surfaces. Prefer browser:control-in-app-browser for unauthenticated browser/page QA in Codex, then Playwright/agent-browser/dev-browser. Captures screenshot/TUI evidence with bundled diff scripts, runs design-system/functional and visual-fidelity/CJK reviewer passes, then synthesizes a good/bad verdict. Triggers: visual QA, screenshot/pixel diff, UI looks wrong, reference fidelity, design system check, responsive check, CJK text clipping, TUI alignment, box-drawing drift, PDF page check, deck review, blank or near-empty page, wrong page break.
- ▌ Dag Library · robertoatilaStore a dag definition once and re-run it in one or two lines, instead of pasting the full definition JSON into every eval cell. MUST USE whenever the user wants to save a DAG for reuse, run a previously saved/named DAG, schedule the same graph repeatedly (nightly/weekly audits, recurring multi-agent pipelines), or asks where to put a dag definition file. Triggers: dag library, save this dag, reuse a dag, run the saved dag, stored dag definition, recurring dag, nightly dag, dag 정의 저장, 저장된 dag 실행, dag 반복 실행, DAG 만들어두고 여러 번.
- ▌
- ▌ Give Me Tips · robertoatilaExplains any senpi tip in depth - startup tips, working tips, and any Tip: line shown in the TUI (including the Fable-5-refusal fallback tip). Use when the user asks about a Tip: line, says give-me-tips, asks what a tip means, how a tipped feature works, or which tips they can see. Queries the live tip list (senpi --list-tips) first, checks what THIS user can actually see, then verifies the real feature code before explaining.
- ▌ Layout Skill · robertoatilaLayer A layout-mechanics reference. Stacks on any style skill when the screen is an app shell, dashboard, settings, list-detail, mail/inbox, or any layout with fixed regions plus a scrolling body — or when a layout breaks under long, empty, or unbroken content. Owns spatial structure and scroll ownership; owns zero visual taste. Load it alongside a style skill; it does not replace one.
- ▌
- ▌ Review Work · robertoatilaPost-implementation review orchestrator. Launches 5 parallel background sub-agents: Oracle (goal/constraint verification), Oracle (code quality), Oracle (security), unspecified-high (hands-on QA execution), unspecified-high (context mining from GitHub/git/Slack/Notion). All must pass for review to pass. MUST USE before a PR handoff or when the user explicitly asks to review completed work. Triggers: 'review work', 'review my work', 'review changes', 'QA my work', 'verify implementation', 'check my work', 'validate changes', 'post-implementation review'.
- ▌ Ultimate Browsing · robertoatilaEscalation skill for blocked or hard-to-reach web access — load it when a normal browse/fetch is blocked (WAF, 403, Cloudflare, JS-only render, login-gated, or a platform a generic fetcher cannot read). Tiered router: TIER 1 insane-search (headless extraction + WAF bypass via curl_cffi TLS impersonation, yt-dlp, Jina Reader, public APIs, Playwright real-Chrome fallback); TIER 1.5 agent-reach (platform-native readers for Chinese and social platforms: Xiaohongshu, Douyin, Weibo, Bilibili, V2EX, WeChat, plus Twitter/Reddit/LinkedIn/GitHub); TIER 2 Chrome stealth (CloakBrowser stealth Chromium + agent-browser CDP for clicks, forms, screenshots, video, cookie login). Triggers: blocked site, bypass bot detection, cloudflare/WAF bypass, scrape, stealth browser, import cookies, fill form, screenshot, play youtube, xiaohongshu, douyin, weibo, bilibili, v2ex, wechat article, podcast transcript. NOT for simple searches (use web-search) or plain fetches (use webfetch).
- ▌ Ultrawork · robertoatilaBinding ultrawork mode directive for omo-senpi. When a prompt contains ultrawork or ulw, the omo input hook injects the full directive as a hidden custom message (customType omo-ultrawork:directive, display false) ahead of the user's text, which is left untouched; a prompt queued while the agent is streaming instead carries the directive appended inside that same message. The directive is present in the conversation context; on the idle path it is not shown in the visible prompt, while a queued prompt carries the directive visibly (exactly as before this change). When the directive is already present in the conversation, do not read this file again - this file is that same directive. Read this file only when ultrawork mode is requested and the directive is not already present in the conversation.
- ▌ Mass Ulw · robertoatilaRun a dependency graph of child agents in one call with the native workflow tool. Use when the user asks for mass-ulw, a DAG of tasks, fan-out/fan-in work, or multi-agent execution where some tasks must wait on others.
- ▌ Pytorch Fsdp2 Per Parameter Sharding · robertoatilaDistributed PyTorch training with per-parameter sharding, mixed precision, and CPU offload.
- ▌ Qformer Multimodal Feature Alignment · robertoatilaBridges frozen vision encoders and LLMs for multimodal image captioning and reasoning.
- ▌ Tensorrt LLM Gpu Kernel Acceleration · robertoatilaAccelerates LLM GPU inference throughput with NVIDIA TensorRT-LLM optimized kernels.
- ▌ Audio Transcription Speech Processing · robertoatilaTranscribes, translates, and processes multilingual speech and audio using OpenAI Whisper models.
- ▌ Biomedical Literature Semantic Search · robertoatilaSearches biomedical literature and extracts structured experimental data via BGPT.
- ▌ Legacy Codebase Reverse Documentation · robertoatilaReverse-engineers legacy codebases to produce missing architectural specs and diagrams.
- ▌ Clip Multimodal Contrastive Embeddings · robertoatilaZero-shot image classification, image-text matching, and multimodal embeddings using CLIP.
- ▌ Guidance Interleaved Token Acceleration · robertoatilaConstrains LLM outputs with grammars and regex for structured data generation.
- ▌ Phoenix Opentelemetry LLM Observability · robertoatilaOpen-source LLM observability, distributed tracing, and evaluation platform using OpenTelemetry.
- ▌
- ▌ Particle Image Velocimetry Fluid Dynamics · robertoatilaExtracts velocity fields and cross-correlation analysis from Particle Image Velocimetry data.
- ▌
- ▌ Verl Hybrid Engine Reinforcement Learning · robertoatilaReinforcement learning framework (PPO, GRPO) for post-training large language models.
- ▌
- ▌
- ▌ Autogen · robertoatilaMulti-agent conversational orchestration framework for autonomous collaborative LLM task execution.
- ▌
- ▌
- ▌
- ▌ Doubt Driven Development · robertoatilaAdversarial code review and verification of architectural decisions before commit.
- ▌ Flaky Test Investigation · robertoatilaDetects and quarantines flaky tests through test history and CI log analysis.
- ▌ Governed Package Publish · robertoatilaSafely publishes packages to npm, PyPI, and Crates.io with two-phase verification.
- ▌ LLM Redteam Plugin Audit · robertoatilaCreates adversarial red-teaming plugins, prompt injection tests, and safety graders.
- ▌ Obsidian Markdown Syntax · robertoatilaCreates Obsidian-flavored markdown notes with wikilinks, callouts, and frontmatter properties.
- ▌ Persistent File Planning · robertoatilaMaintains persistent task_plan.md, findings.md, and progress.md files across agent sessions.
- ▌