183 skills · sorted by relevance
Break an RFC, a written design proposal, into work units with dependencies and acceptance tests for multiple agents. Organize research, implementation, testing, review, queued merges, and final system verification for features too large for one pass.
Guide market prediction workflows with LSTM, Transformer, and N-BEATS models using neural-trader commands and result storage. It explicitly identifies version 2.8.11 outputs as placeholders, so they cannot be treated as fitted models or evidence of prediction accuracy.
Take implementation work through small, independently mergeable pull requests. Use a separate working directory for each PR, record manual QA evidence, write English review descriptions, resolve CI and review findings, then merge and clean up.
Breaks long-running goals into small, verifiable steps, tracks progress, and records evidence from actual operation. Checkpoints preserve what has been proven so work can resume after interruptions and continue toward explicit completion criteria.
Design AI agents that break down goals, plan actions, use tools, and correct their own mistakes. The guidance focuses on reliable execution, clear operating boundaries, and managing conversation context as tasks grow longer.
Organize complex work for a team of AI agents by splitting tasks, defining dependencies, and documenting ownership and completion criteria. Provides strategies for tracking progress, spotting blockers, and redistributing work when workloads become uneven.
Guide a development brief through specification, dependent tickets, implementation, review, and verification. Coordinate parallel work while keeping human approval at key decisions and recording agreed vocabulary, business rules, and the reasons behind technical choices.
Break down major industrial trends into materials, components, and equipment to find listed suppliers at supply bottlenecks. Compare scarcity, financial performance, and valuation, then organize candidates into a ranked research board with supporting evidence and risks.
Use SAELens to train and analyze sparse autoencoders, which separate neural network activity into features that may represent recognizable concepts. Explore pretrained features, evaluate reconstruction quality, and investigate how individual features influence language model outputs.
Design autonomous AI agents that break goals into tasks, use tools, and review their own work. Covers reasoning and action loops, separate planning and execution, and reliability measures such as cost limits, validation, and logging.
Analyze differences between financial budgets, actual results, forecasts, or prior periods. Break changes into measurable drivers, explain business causes, and prepare management commentary with waterfall breakdowns that show how each factor contributes.
Build three-tier bookings forecasts from sales opportunities and historical conversion data. Project recurring revenue by customer cohort, assess the reliability of each sales stage, and present forecast numbers alongside their assumptions.
Break a game concept into the systems it needs, including supporting systems that are easy to overlook. Map dependencies, assign priorities by development stage, and create an index showing the recommended order for detailed design.
Stress-test ideas, strategies, and plans by breaking them into individual claims for parallel expert agents to challenge. Receive the strongest supporting case, surviving counterarguments, and findings ranked by severity with suggested ways to address weaknesses.
Turn a rough research approach into a focused plan with concrete implementation details and a small set of validation experiments. Repeated Claude reviews help refine the method while keeping the original problem and resource constraints in view.
Turn a vague research approach into a concrete method proposal with focused contributions and minimal validation experiments. Ground the plan in relevant papers, then refine it through GPT-6-Astra reviews while preserving the original problem and resource constraints.
Develop a vague research approach into a concrete method proposal through repeated Gemini reviews. Keep the original problem and constraints in focus while specifying implementation details, removing unnecessary components, and outlining the minimum experiments needed to test the claims.
Use SAELens to train and analyze sparse autoencoders, which break model activations into a small set of interpretable features. Explore pretrained features, evaluate reconstruction quality, troubleshoot training, and test how individual features influence generated text.
Design systems made of independently deployable services, or plan how to split an existing application. The skill produces service boundary diagrams and recommendations for communication, data ownership, failure handling, monitoring, and deployment.
Break a larger goal into two to six tasks and assign a worker agent to each in sequence. Check results with shell commands, retry failed work, and report completion using the configured default model.
Decompose large tasks across parallel agents — use for migrations, multi-file refactors, or batch work
The English summary is being prepared. Showing the original description.
Decompose and execute large changes, migrations, or multi-issue fixes in parallel with quality gates
The English summary is being prepared. Showing the original description.
矛盾分析法：把复杂问题拆成若干对立面，找出规定其他矛盾的主要矛盾及其主要方面，判定对抗性 / 非对抗性，并据此选择处理方式。当问题头绪多、多个因素互相牵制、优先级不清、根因不明、反复修不好、trade-off 说不清时触发；直接执行类任务或用户已定方案时不触发。 English: Contradiction analysis. Decompose a tangled problem into opposing forces, isolate the principal contradiction and its dominant side, classify it as antagonistic or non-antagonistic, and choose the response accordingly. Trigger on unclear priorities, unknown root causes, recurring failures, or trade-offs that resist explanation; skip for direct execution or when the user has already chosen the solution.
The English summary is being prepared. Showing the original description.
A goal-like loop that decomposes work into systematic, evidence-bound ultrawork steps. Use when the user wants a goal loop or durable, checkpointed execution.
The English summary is being prepared. Showing the original description.