Use when you have a hunch, a candidate optimization, a novel technique, or an "I think X is faster/better/possible" idea for tensor-grep and must turn it into an ACCEPTED result…
Visualize training metrics, debug models with histograms, compare experiments, visualize model graphs, and profile performance with TensorBoard - Google's ML visualizatio — from…
Expert guidance for ClearML, the open-source MLOps platform for experiment tracking, pipeline orchestration, data management, and model deployment.
Expert guidance for Comet ML, the platform for tracking machine learning experiments, managing models, and monitoring production ML systems.
Expert guidance for Malloy, the experimental data language from Google that replaces SQL for analytics with a composable, reusable, and more readable syntax.
Expert guidance for user research, helping product teams conduct interviews, create personas, design surveys, run usability tests, and synthesize findings into actionable…
Termination review — high-risk flag detection, severance + release, and final pay timing by jurisdiction.
Termination review — automatic unfair dismissal flag detection, notice and final pay, statutory redundancy (where applicable), ACAS Code compliance, and Settlement Agreement gate.
Termination review — high-risk flag detection, severance + release, and final pay timing by jurisdiction.
Use when a solution needs a coherent structure before implementation or production for Software testing and release assurance: unit testing for cross-platform behavior.
Use for LaTeX maths and references — amsmath, theorems, cleveref, biblatex, fonts, algorithms.
Computational text analysis for sociology research using R or Python. Guides you through topic models, sentiment analysis, classification, and embeddings with systematic…
LLM-based text classification. Codebook, validation, agreement statistics.
A local text formatting organizer for biomedical/academic writing; use it when you need to clean whitespace/line endings while preserving Markdown structures or when normalizing…
textstat is a Python package that calculates readability statistics for text content, including Flesch Reading Ease, Gunning Fog Index, SMOG Index, Coleman-Liau Index, and…
Use when targeting The Astronomical Journal (AJ) or deciding whether an observational astronomy manuscript fits this AAS/IOP venue.
An experimental C++20 browser engine derived from Ladybird/SerenityOS, built with CMake + vcpkg and tested with CTest; explicitly not production-ready and honest about provenance.
Use when targeting The EMBO Journal (EMBO J) or deciding whether a mechanistic molecular biology manuscript fits this venue.
Use when targeting The International Journal of Robotics Research (IJRR) or deciding whether a robotics manuscript fits this venue.
Use when a question, decision, plan, tradeoff, or claim needs a rigorous verdict and one perspective is not enough.
Use when targeting The Lancet Diabetes & Endocrinology or deciding whether a diabetes, endocrine, or metabolic study fits this venue.
Use when targeting The Lancet Neurology or deciding whether a clinical neurology manuscript fits this venue.
Master AI-native product orchestrator. Cultivates living intelligence products by weaving capabilities from every domain (design, data science, investing, writing, game theory,…
Use when the user asks to "use the oracle" or "ask the oracle" for deep research, analysis, or architectural questions requiring multi-source investigation.
Use when targeting The Plant Cell or deciding whether a plant biology manuscript fits this venue. Encodes the journal's fit, framing, method-and-evidence bar, house style,…
Use when targeting The Plant Journal or deciding whether a plant-science manuscript fits this venue. Encodes the journal's fit, the mechanistic gene-to-function and…
Research, compare, sample, select, persist, adapt, rotate, and archive reusable visual themes and design templates
Use when performing mathematical physics computations - Lie algebras, quantum chemistry, neural operators, theorem proving, or scientific validation.
Use when when you have peak-abundance .csv files with assigned molecular formulas (elemental composition: C, H, O, N, P, S) from FT-ICR MS or high-resolution MS and need to…
Thermodynamics from the zeroth law through statistical mechanics. Covers temperature and thermal equilibrium, heat transfer mechanisms (conduction, convection, radiation), the…
Build a defensible thesis or dissertation outline — argument-first structure, chapter by chapter, with the through-line visible.
Combined research, decision, and codification workflow. Use when: (1) Exploring a question before committing (2) Making a decision that needs documentation (3) User says — from…
Analysis of Competing Hypotheses. Operationalizes Richards Heuer's CIA-tradition technique for systematically narrowing among multiple hypotheses against evidence.
Produces a clustered theme map that groups many raw notes, observations, quotes, or data points bottom-up into a small set of named, traceable themes (the KJ method).
Complex ideas need iteration, not conversation. Think → Do → Review → Repeat. CellCog completes the full loop — brainstorming, reasoning, problem-solving, strategic thinking, then…
Abductive reasoning about causes. Takes a phenomenon, separates observations from interpretations, then spawns diagnosticians applying different reasoning lenses (technical,…
Apply Karpathy framing — AI agent 架構 / skill as code / auto-research / 學習工作流 / human-vs-agent split / 教育設計.
Researches a thinking framework end to end and produces a schema-valid proposed registry entry plus a learning dossier, grading the evidence on the seven-tier model and assessing…
First-person rigor patterns for non-trivial work — type/dimension check, decompose, bound-then-refine, sanity at limits, falsifiability.
Use when adding a new policy-heavy specialist agent, when reviewing why an existing agent is making sloppy mid-flight decisions, or when the model is acting on tool output without…
Multi-mode analytical thinking — first principles decomposition, iterative depth analysis, multi-agent council debates, adversarial red teaming, and scientific hypothesis testing.
Apply parsimony principle to prefer simpler explanations with fewer assumptions. Use for hypothesis selection in debugging, architecture decisions, and choosing between competing…
Multi-source research on concepts and ideas. Combines web search, documentation lookup, and codebase analysis into structured findings.
Hypothesis → Prediction → Test → Revise with explicit falsification. Use for debugging, feature experimentation, performance investigation, and A/B testing design.
Test ideas through hypothetical scenarios when empirical testing is impractical. Use for architecture evaluation, edge case analysis, ethics considerations, and strategy…
Use when tackling complex reasoning tasks requiring step-by-step logic, multi-step arithmetic, commonsense reasoning, symbolic manipulation, or problems where simple prompting…
Activate when: user says 'what if,' 'let's reason through this,' 'imagine that,' or 'thought experiment'; a strategic decision is too risky or irreversible to test empirically;…
Construct, analyze, and systematically vary thought experiments to test intuitions and principles. Use when the user needs to test whether a principle holds under extreme…
Run a dynamic ThreadOps research workflow for current-doc lookup, technical comparison, architecture options, cross-checked source synthesis, or decision support.
Three-layer skill quality evaluation (Static → LLM Judge → Monte Carlo). Layer 1: deterministic checks (frontmatter, triggers, size, links) in <2s.
Implement browser-game runtimes with plain Three.js. Use when the user wants imperative scene control in TypeScript or Vite with GLB assets, loaders, physics, and low-lev — from…
Design a collision and physics layer for a three.js game, including character movement, collision layers, scene queries, and engine-adapter boundaries.
Build and iterate playable Three.js game systems. Combines starter scaffold creation, architecture, gameplay implementation, and game-feel tuning.
Use when you have downloaded or cloned a fragmentation library repository (such as LipidMatch) and need to verify that it contains the expected breadth of coverage across both…
Comprehensive guide for developing MailExtensions for Mozilla Thunderbird, including Manifest V2/V3 configuration, all messenger.* APIs, UI actions, Experiment APIs, and ATN…
NotebookLM CLI wrapper via `node {baseDir}/scripts/notebooklm.mjs`. Use for auth, notebooks, chat, sources, notes, sharing, research, and artifact generation/download.
Contract-driven ticket execution with Linear integration - orchestrates intake, research, questions, spec, implementation, review, and done phases with explicit human gates
Use when a task needs the judgment of a Tile and Stone Setter — selecting a shower waterproofing method against the drain type, specifying mortar coverage and trowel size for…
Exhaustive problem space exploration using the MIT Synthetic Neurobiology "tiling tree" method. Partitions a problem into MECE (Mutually Exclusive, Collectively Exhaustive)…
Time series analysis and forecasting methods from classical statistics to foundation models. Reference when decomposing temporal patterns, selecting forecasting models, handling…