「refinement」の検索結果

Polishing pass over all open beads. Run 4–6 times. "Check your beads N times, implement once." Use after plan-to-beads and between refinement rounds.

日本語の概要は準備中です。原文の説明を表示しています。

Zod schema validation patterns for TypeScript - schema definitions, type inference, refinements, transforms, discriminated unions

日本語の概要は準備中です。原文の説明を表示しています。

N-round adversarial review pipeline for empirical research output — the chain from data to LaTeX tables to a manuscript that cites them. A Claude drafter proposes minimal diffs, a deterministic mechanical battery gates every diff from a clean state with a regression gate, a Codex reviewer files check-backed critiques, and a blind judge panel decides residual disputes. Manual-invoke ONLY: trigger when the user explicitly runs /adversarial-empirical-review or names 'adversarial-empirical-review' / 'adversarial empirical review'. Do NOT auto-trigger on generic 'review my results', 'check my tables', or manuscript-editing requests. For prose-style refinement use style-emulation instead; this skill AUDITS WHETHER THE TABLES ARE CORRECT — that each number in the tables is what the analysis code computes, reproduces from the data, and is internally consistent. It is an empirical + code review: the manuscript is read only to resolve table numbering, and prose is not examined.

日本語の概要は準備中です。原文の説明を表示しています。

Craft effective AI image prompts with structured syntax, style control, and iterative refinement techniques.

日本語の概要は準備中です。原文の説明を表示しています。

Use this skill when designing, generating an MVP blueprint for, auditing, troubleshooting, refactoring, or explaining an agentic harness for any domain. Covers provider-neutral agent architecture for OpenAI, Anthropic, and OpenAI-compatible APIs: agent loops, tool design, record provenance, interactive presentation, user-memory lifecycles, environment-adaptive tools, speculative tool execution, late-bound capabilities, permissions, system prompts, planning, goals, always-on agents and durable runtime, adaptive agent teams, context compaction, memory, skills, MCP/external connectors, public-board communications, hardware agents and board deployment, self-refining recursive harnesses, programmable context, continual refinement, observability, evals, prompt caching, agent-legible environments, feedback loops, and safety.

日本語の概要は準備中です。原文の説明を表示しています。

Refine an existing implementation plan with a second iteration. Re-analyzes the codebase for gaps, missing tasks, and wrong dependencies. Use after /unikit-plan or to improve a /unikit-fix plan, or when the user says "improve the plan", "what's wrong with the plan", or "update the plan from research". Optional +check flag validates refinements via a fresh-context subagent.

日本語の概要は準備中です。原文の説明を表示しています。

Manages intent — features, enhancements, bugs and the reasoning behind them — recorded as prose in the knowledge base with a lifecycle. Use when capturing new work, moving work through backlog, refinement or implementation, releasing or cancelling work, or asking what is pending, in flight, or no longer valid.

日本語の概要は準備中です。原文の説明を表示しています。

Refine academic writing for computer science research papers targeting top-tier venues (NeurIPS, ICLR, ICML, AAAI, IJCAI, ACL, EMNLP, NAACL, CVPR, WWW, KDD, SIGIR, CIKM, and similar). Use this skill whenever a user asks to improve, polish, refine, edit, or proofread academic or research writing — including paper drafts, abstracts, introductions, related work sections, methodology descriptions, experiment write-ups, or conclusion sections. Also trigger when users paste LaTeX content and ask for writing help, mention "camera-ready", "rebuttal", "paper revision", or reference any academic venue or conference. This skill handles both full paper refinement and section-by-section editing.

日本語の概要は準備中です。原文の説明を表示しています。

Take a chosen visual to finished, production-ready state. Runs cleanup + refinement passes (fix artifacts, balance composition, color-grade, sharpen text, conform to spec/aspect) using image edit/regenerate. Use when the user says "polish this", "finalize", "clean up the image", "production-ready", "fix the artifacts", or has a selected visual that needs finishing. Final stage of the Creative Production set.

日本語の概要は準備中です。原文の説明を表示しています。

Use when user says "simmer this", "refine this", "hone this", "iterate on this", or asks to improve a specific artifact over multiple rounds. Runs an iterative refinement loop with investigation-first judges that read the code, understand the problem, and propose evidence-based improvements. Auto-selects single judge or multi-judge board based on complexity. Works on any artifact type: documents, prompts, specs, emails, creative writing, API designs, pipelines, codebases. Supports multi-file workspace targets, runnable evaluators, and open-ended optimization (model selection, pipeline topology, prompt tuning).

日本語の概要は準備中です。原文の説明を表示しています。

Generates 3D conformer ensembles using RDKit ETKDGv3 with knowledge-enhanced distance geometry, MMFF94/UFF force-field optimization, CREST + GFN2-xTB semi-empirical refinement, and macrocycle-aware torsion preferences. Provides explicit decision rules for single vs ensemble conformer use, RMSD pruning, energy windows, conformer count, and force-field choice. Use when preparing 3D ligands for docking, generating descriptor input for 3D QSAR, or sampling macrocycle/peptide conformational ensembles.

日本語の概要は準備中です。原文の説明を表示しています。

Process and generate multimedia content using Google Gemini API. Capabilities include analyze audio files (transcription with timestamps, summarization, speech understanding, music/sound analysis up to 9.5 hours), understand images (captioning, object detection, OCR, visual Q&A, segmentation), process videos (scene detection, Q&A, temporal analysis, YouTube URLs, up to 6 hours), extract from documents (PDF tables, forms, charts, diagrams, multi-page), generate images (text-to-image, editing, composition, refinement). Use when working with audio/video files, analyzing images or screenshots, processing PDF documents, extracting structured data from media, creating images from text prompts, or implementing multimodal AI features. Supports multiple models (Gemini 2.5/2.0) with context windows up to 2M tokens.

日本語の概要は準備中です。原文の説明を表示しています。

Autonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.

日本語の概要は準備中です。原文の説明を表示しています。

Identifies when task outputs require iteration based on quality signals, unmet requirements, or explicit feedback. Triggers appropriate re-execution strategies. Activate on 'needs iteration', 'retry needed', 'not good enough', 'try again', 'refine output'. NOT for feedback generation (use dag-feedback-synthesizer) or convergence tracking (use dag-convergence-monitor).

日本語の概要は準備中です。原文の説明を表示しています。

Architect systems like Fable 5 — deep understanding before design, modular thinking, iterative refinement. Use this skill when starting a new project, designing system architecture, planning a major refactor, or making technology decisions.

日本語の概要は準備中です。原文の説明を表示しています。

[Project Management] Use when capturing new ideas, feature requests, or concepts for future refinement.

日本語の概要は準備中です。原文の説明を表示しています。

Solves complex, open-ended problems using a knockout tournament mechanism. Generates an initial population of diverse candidate answers, pairs them into head-to-head matches, produces a hybrid "child" solution from each pair, and advances the best of the three through a bracket until a winner emerges. Use this skill whenever the user wants to competitively select the best answer from a field of alternatives, asks for a "tournament" or "bracket" style search, wants a faster alternative to full evolutionary search, or uses phrases like: "run a tournament on this", "bracket-style search", "let solutions compete", "knockout search", "find the best answer by elimination", or "which solution wins?". Also trigger when the problem is open-ended and the user wants a clear winner quickly rather than a deep multi-generation refinement.

日本語の概要は準備中です。原文の説明を表示しています。

Use when work needs Scrum roles, ceremonies, backlog refinement, sprint planning, story delivery, release readiness, or role handoffs.

日本語の概要は準備中です。原文の説明を表示しています。

Reconstruct an evidenced indoor space from local Realsee scan exports, point clouds, CAD, and panoramas as a separately editable native Blender scene. Use for scan-to-Blender space modeling, source-based geometry/material/light refinement, and requested walkthrough or physics exports. Argus panorama-to-depth inference is a separate capability.

日本語の概要は準備中です。原文の説明を表示しています。

Load for Senior CTO critique, architectural audit, risk review, sequencing correction, or refinement of an existing implementation-plan triad; edits the requested plan/context/tasks in place without creating a worktree or review file. Not for open-ended CTO advice, product implementation, or implementation workspace setup.

日本語の概要は準備中です。原文の説明を表示しています。

Knowledge base from the Expanded Guidance for NASA Systems Engineering (NASA/SP-2016-6105-SUPPL, Volume 1, March 2016) — the practitioner-depth supplement to the NASA SE Handbook. Use for the load-bearing mechanics the handbook compresses: the SE Engine's per-phase cadence and the iterative-vs-recursive distinction, the AS9100 crosswalk, life-cycle tailoring vs. customization and the Compliance Matrix, the recursive system-design consistency loop (ConOps vs. Operations Concept, requirement flow/type/ownership, successive refinement), product realization (verify/qualify/accept/certify cardinality, protoflight, 'test the way we fly' coverage, two-form transition), and the crosscutting technical-management procedures (six-step scheduling, cost-estimating method selection, JCL, interface document family, RIDM/CRM, CM baselines, EVM arithmetic, MOE/MOP/KPP/TPM, the three required leading indicators, and the decision-analysis method catalog). This is the DEPTH layer — it deliberately does NOT restate the base definitions; pair it with the nasa-se-handbook pack. Scope: NASA space-flight SE practice (Vol 1 Practices); thin on the Vol 2 process-appendix detail, on non-NASA/commercial life cycles, and on software-engineering methodology.

日本語の概要は準備中です。原文の説明を表示しています。

Add or define specific fields in a collection's readSchema. Use when fields are missing from materialization field selection, pruned by complexity limit, need type refinement, need to be protected from future inference changes, or need to be required for downstream consumers. Use when user says "missing fields", "pruned fields", "field not showing up", "complexity limit", "define field in read schema", "restrict field type", or "protect field from inference".

日本語の概要は準備中です。原文の説明を表示しています。

Autonomously improve a real artifact (code, training recipe, agent harness, data pipeline, prompt) against an objective and an evaluator, using Hypothesis Tree Refinement (HTR) from the Arbor paper. Use this whenever someone wants to iteratively optimize something over many experiments without overfitting — e.g. "get my model's eval score up", "improve this agent/harness", "tune this pipeline", "beat the baseline on this benchmark", "run a search over approaches and keep the best", "do an MLE-bench / Kaggle-style optimization", or any long-horizon "make this artifact better and don't just memorize the dev set" task. Trigger it even when the user doesn't say "Arbor" or "hypothesis tree" but describes repeated experiment-and-evaluate loops, branching exploration of competing ideas, or worries about a dev/test gap. Runs Claude itself as the coordinator with subagent executors in isolated git worktrees; for the standalone `arbor` CLI tool see references/arbor-upstream.md.

日本語の概要は準備中です。原文の説明を表示しています。

Interactive idea refinement using Socratic method to develop fully-formed designs

日本語の概要は準備中です。原文の説明を表示しています。