「ide」の検索結果

Use when turning something that happened into something that makes the next run better - "write a retro", "record this paper cut", "this bug escaped verification", "make this an eval", "why do we keep hitting this", "did that change actually help", "scorecard", "what did we learn from that run", "lessons keep getting re-learned". Covers the six-step loop (capture, classify, convert, change, measure, retro), the repo artifact each step produces, and how escaped bugs become eval cases. Do NOT use to design or judge a single workflow experiment (see `workflow-experimentation`, which this loop feeds), to decide where to run work (see `dogfooding`), or to debug one failed execution in the moment.

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 acting as the orchestrator agent that drives Syntropic137 - running the dogfood loop, deciding what to dispatch next, landing agent PRs, deploying a beta, triaging failed executions, or keeping the scratchpad, paper cuts and retrospectives current. Trigger phrases include "orchestrator tick", "keep the loop running", "what should we dispatch next", "land these PRs", "why are PRs stacking up", "deploy the next beta", "pit stop", "the run failed, what now", "write the retro", "is the system helping", "scale quality", "learning loop for the orchestrator". Do NOT use for deciding whether one task belongs in a workspace or locally (see `dogfooding`), for turning one incident into an eval case (see `learning-loop`), for designing a workflow A/B experiment (see `workflow-experimentation`), or for release-branch and version-bump mechanics (see `devops`).

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 asked to report the deployment sentinel value. Provides the sentinel string that proves a vendored skill reached the agent's context.

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 deciding WHERE to do work on Syntropic137 itself - "run this through Syntropic", "dispatch a workflow", "should this go in a workspace or local", "kick off a run on the Mini", "dogfood this", "build Syntropic inside Syntropic", "why did the agent's push fail", "can an agent run our QA", "the execution says completed but nothing shipped". Covers what a workspace agent can and cannot do, why `.github/workflows` pushes must come from the local machine, how to write a task prompt that does not repeat a known failure, and how to verify a run actually delivered. Do NOT use for PR/release mechanics (see `devops`), for using the `syn` CLI as an end user (see the syntropic137 plugin skills), or for authoring workflow YAML.

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。

Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

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。

Guide for checkpointing — saving, loading, and resuming training with CheckpointRecord. Use when the user asks about saving weights, resuming training, checkpoint management, or the checkpoint lifecycle.

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。

Guide for testing conventions and CI pipelines — unit tests, integration smoke tests, pytest markers, and GitHub Actions workflows. Use when the user asks about testing, CI, running tests, or adding tests for a recipe.

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。

Guide for using renderers — the bridge between chat-style messages and token sequences. Covers renderer setup, TrainOnWhat, vision inputs, model family matching, and custom renderers. Use when the user asks about renderers, tokenization, message formatting, or vision inputs.

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。

Guide for hyperparameter selection — learning rate formulas, LoRA rank, batch size, group size, schedules, and model-specific tuning. Use when the user asks about learning rate, batch size, hyperparameter tuning, or how to configure training parameters.

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。

Guide for defining RL environments — the Env protocol, EnvGroupBuilder, RLDataset, and custom environment creation. Use when the user asks about RL environments, reward functions, or how to define custom tasks for RL training.

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。

Guide for contributing to the tinker-cookbook repo — development setup, code style, type checking, PR process, and design conventions. Use when the user asks about how to contribute, set up the dev environment, code style, or project conventions.

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。

Guide for the weight lifecycle — downloading trained weights from Tinker, merging LoRA adapters into HuggingFace models, and publishing to HuggingFace Hub. Use when the user asks about exporting, downloading, merging, or publishing trained model weights.

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。

Guide for installing Tinker, setting up the environment, getting an API key, and verifying everything works. Use when the user is getting started, setting up their environment, or troubleshooting installation issues.

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。

Guide for evaluation — inline evaluators, Inspect AI integration, and custom evaluators for measuring training progress. Use when the user asks about evaluation, metrics, benchmarks, or how to measure model quality during training.

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。

Guide for using completers — TokenCompleter and MessageCompleter for text generation during RL rollouts and evaluation. Use when the user asks about generating text, completing messages, or using completers in RL environments.

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。

Guide for training outputs, metrics logging, logtree reports, tracing/profiling, and debugging training runs. Use when the user asks about training logs, metrics, debugging, tracing, profiling, timing, Gantt charts, or understanding training output files.

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。

Guide for using the Tinker Python SDK APIs — ServiceClient, TrainingClient, SamplingClient, RestClient, forward_backward, optim_step, sampling, and async patterns. Use when the user asks about Tinker API basics, how to call training/sampling, or how the SDK works.

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。

Guide for choosing models in Tinker — available model families, model types (base, instruction, reasoning, hybrid, vision), architecture (dense vs MoE), and how to match renderers to models. Use when the user asks which model to use, what models are available, or how to pick a model for their task.

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。

Guide for dataset construction — SupervisedDatasetBuilder, RLDatasetBuilder, ChatDatasetBuilder, and custom dataset creation from JSONL, HuggingFace, or conversation data. Use when the user asks about datasets, data loading, data preparation, or custom data formats.

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。

Guide for the Tinker CLI — managing training runs, checkpoints, downloading weights, and publishing to HuggingFace. Use when the user asks about CLI commands, listing runs, managing checkpoints from the terminal, or uploading to HF Hub.

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: planner mode, planning mode, proposal, scope definition, estimate, timeline, how hard, feasibility, effort, how long, how big, client message, external communication, pricing, delivery confidence, commercial risk, or project delivery planning.

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。

Review Codex/Claude agent skills, profile manifests, and skill-related deploy changes before they ship. Use when creating, editing, importing, installing, or reviewing SKILL.md files, agent manifests, optional third-party skills, or plugin skill bundles. Checks trigger quality, progressive disclosure, evidence gates, security, hallucination risk, validation, and deployment impact.

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 or reviewing cover letters, recruiter email replies, LinkedIn outreach, or interview follow-ups. Removes AI-template patterns and writes with a specific, confident, human voice.

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。

Guides systematic root-cause debugging. Use when tests fail, builds break, behavior doesn't match expectations, or you encounter any unexpected error. Use when you need a systematic approach to finding and fixing the root cause rather than guessing.

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。