「harness」の検索結果

lm-eval-harness: benchmark LLMs (MMLU, GSM8K, etc.).

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 writing, debugging, or configuring an omp hook or extension — blocking or rewriting tool calls, redacting tool output, reshaping LLM context, registering tools/commands, or wiring modules through config.yml. Covers ExtensionAPI (current) vs HookAPI (legacy), the event catalog, discovery and load order, and disable controls. Not for Claude Code hooks (~/.claude/settings.json) or the shared harness safety guard (~/.agents/harness/hooks/).

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。

Improve agent harness configuration based on recent friction points. Use when invoked as `$retro` or when the user asks for autonomous cleanup of Codex, agent harness, or migrated Claude configuration.

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。

Harness-first workflow for evaluating and improving terminal UIs through real interaction. Use this whenever the user wants to review a TUI experience, compare before/after behavior, identify UX friction, validate navigation or filtering flows, inspect visual stability, or turn observed interaction problems into concrete follow-up tests and improvement recommendations. Prefer this skill any time a TUI should be exercised through a reproducible PTY/session harness instead of only reading code.

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。

Avalia LLMs em mais de 60 benchmarks acadêmicos (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use quando benchmarking de qualidade de modelo, comparação de modelos, relatório de resultados acadêmicos ou rastreamento de progresso de treinamento. Padrão da indústria usado por EleutherAI, HuggingFace e laboratórios principais. Suporta HuggingFace, vLLM, APIs.

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。

[Quality] Use when setting up an agent quality harness with feedforward guides and feedback sensors.

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。

Patient safety evaluation harness for healthcare application deployments. Automated test suites for CDSS accuracy, PHI exposure, clinical workflow integrity, and integration compliance. Blocks deployments on safety failures. Use when a healthcare deployment must be gated on patient-safety tests for CDSS accuracy, PHI exposure, and workflow integrity.

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。

How to MEASURE this repo instead of eyeballing it — interpret the loop-station context-budget trend (history.jsonl, thresholds, the tokenizer-switch gotcha), read a session trace, use the eval ledger (cost-per-success), and pick the right evidence tool (app-shots, smoke-deployed, db-counts, --check modes). Use when asked "is the harness bloating", "how big is CLAUDE.md really", "did this agent run actually succeed", "prove this works", "what evidence do we have", or when tempted to declare a capability (browser, tool) unavailable. For RUNNING the loop-station scripts, defer to the loop-station skill — this skill is interpretation + the wider evidence toolkit.

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。

Observe the harness itself — (WS1) measure context budget as a deterministic, trendable inventory (tokens per CLAUDE.md / skill / agent, lexical redundancy, per-CI cold-write totals; committed history.jsonl), (WS2) reconstruct the real invocation trace (which skills/agents/tools fired and how they nested) from session transcripts, and (WS-A) roll the per-run CI traces up into a committed weekly usage record (per-skill invocation counts; usage.jsonl) — the usage side of the dead-weight join. Use for "context budget", "how big is CLAUDE.md / our skills", "is the harness bloating", "how do our agent loops actually run", "which skills actually get used", "trace this session", "loop station".

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。

Central entry point of the afonsoft agent harness. Use when starting a new project, resuming an existing one, planning features/Epics/releases, or running any multi-step agent-driven work. Validates and reconciles SPECs (SDD), audits the codebase and harness for gaps (security, architecture, performance, hygiene), proposes improvements, fragments work into GitHub Issues, delegates implementation/QA/review to specialized skills, and re-validates everything until delivery. Also use to review unapproved SPECs, reconcile open GitHub Issues with code, or run a final gap check before closing a release.

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 initializing or migrating an AI agent harness in a repository.

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。

【穴埋め雛形】このプロジェクト固有の「存在設計・体験設計」のSSOT。judgment-harness のプロジェクト層スキルとして、立ち上げ時に本ファイルを埋めて使う（雛形のままなら未確定＝推測で決めない）。デザイン/UI/モーション/コピーの方針判断、見た目や体験のトーン決め、画面・コンポーネント設計の前段で必ず参照する。

Run or change the macOS UX trace harness for action-to-visual-completion latency, scroll stability, fixture pressure, and baseline evidence.

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。

Orchestrate any tdmcp test-coverage work: run the coverage harness, inspect coverage gaps, plan focused Vitest/msw or bridge tests, delegate coverage writers, verify coverage deltas, and re-run gates. Use whenever the user asks to raise coverage, add broad regression coverage, create a test harness, improve tests, re-run a coverage wave, or fix a coverage gate.

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。

lm-eval-harness: benchmark LLMs (MMLU, GSM8K, etc.).

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。

Study a merged or shipped tdmcp implementation and extract reusable learnings: code improvements, runtime/UX lessons, test gaps, docs updates, roadmap items, and harness changes. Use when the user asks to learn from a completed feature/project/PR/build, analyze what can be improved from an implementation, or turn a real installation experience into actionable tdmcp improvements. This is a post-implementation learning harness; it does not replace tdmcp-pipeline for building or tdmcp-quality-audit for broad repo audits.

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。

Configure and synchronize multi-harness subagents (Claude Code, Google Antigravity, OpenCode, ChatGPT Codex CLI) with optional per-harness model overrides and single source of truth in .aos/pipeline.json.

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。

Delegate a task to another installed CLI harness (Antigravity/agy, OpenCode, Codex) from inside an AOS repo. Use this instead of shelling out to that CLI directly whenever one is installed — it wraps the wrapper flags for you and streams live tool-call telemetry to agenttrail, which is why the board updates in real time. Reach for it any time you're about to delegate work to another harness.

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。

Harness patterns for coding agents — memory, permissions, context engineering, delegation, skills, hooks, bootstrap.

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。

Evaluates LLMs across 60+ academic benchmarks (MMLU, HumanEval, GSM8K, TruthfulQA, HellaSwag). Use when benchmarking model quality, comparing models, reporting academic results, or tracking training progress. Industry standard used by EleutherAI, HuggingFace, and major labs. Supports HuggingFace, vLLM, APIs.

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。

Evolve this harness with Darwin Mode — frozen model, evolving harness (real, sandboxed, safety-gated).

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。

Always use browser-harness for any web interaction: automation, scraping, testing, or site/app work.

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。

MAGI System v2 — AoT 分解 + MELCHIOR/BALTHASAR/CASPAR 合議 + lam-harness:gabriel adversarial probe による 構造化意思決定フレームワーク。判断ポイント 2+ / 影響レイヤー 3+ / 選択肢 3+ で使用。 Use when facing complex decisions with multiple trade-offs or architectural choices.