メニュー
Browse, promote, demote, classify, and generate training examples from Claude Code execution traces. Use when asked about traces, gold traces, trace counts, promoting or demoting traces, generating examples, exporting training data, or syncing traces.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
Read and control I2C and SPI peripherals on Sipeed boards (LicheeRV Nano, MaixCAM, NanoKVM).
Submit development specs, approve execution plans, monitor workers, retry failed tasks, check job status, and manage orchestration jobs. Use when asked to start a development job, check worker progress, approve a plan, retry something, or list orchestrator jobs.
Generate synthetic training data, manage seeds, preview synthesis jobs, and work with the data factory. Use when asked to generate synthetic data, create training examples, augment a dataset, manage seeds, or check factory job status.
Browse, compare, evaluate, and manage trained models. View leaderboard, trends, lineage, and artifacts. Deploy models to Ollama. Use when asked about models, model comparison, which model is best, model evaluation, deploying a model, or downloading a model.
Check BashGym system status, GPU utilization, hardware info, and aggregate statistics. Use when asked about system health, what's running, GPU usage, memory, disk space, or general system stats.
Start, stop, pause, resume, and monitor model training runs. Check training progress, view metrics, and get GPU utilization during training. Use when asked to train a model, check training status, view loss curves, stop a run, or see training logs.
| name | traces |
| description | Browse, promote, demote, classify, and generate training examples from Claude Code execution traces. Use when asked about traces, gold traces, trace counts, promoting or demoting traces, generating examples, exporting training data, or syncing traces. |
Manage Claude Code execution traces and convert them to training examples.
Filter by status (gold, pending, failed):
scripts/api.sh GET "/traces?status=gold"
scripts/api.sh GET "/traces?status=pending"
scripts/api.sh GET /traces
Quick counts by status:
scripts/api.sh GET /traces/stats
See which repos have traces:
scripts/api.sh GET /traces/repos
scripts/api.sh GET /traces/gold
Move a pending trace to gold status:
scripts/api.sh POST /traces/{trace_id}/promote '{}'
Destructive — confirm with user first. Move a gold trace back to failed:
scripts/api.sh POST /traces/{trace_id}/demote '{}'
Convert a trace into training examples:
scripts/api.sh POST /traces/{trace_id}/generate-examples '{}'
scripts/api.sh GET /training/examples
Export generated examples for training:
scripts/api.sh POST /training/export '{}'
Import new traces from Claude Code history:
scripts/api.sh POST /traces/sync '{}'
Automatically classify pending traces:
scripts/api.sh POST /traces/auto-classify '{}'
User: "How many gold traces do I have?" → Call GET /traces/stats. Report counts per status.
User: "Promote trace abc123" → Call POST /traces/abc123/promote. Confirm success.
User: "Generate training data from my gold traces" → Call GET /traces/gold to list them. For each, call POST /traces/{id}/generate-examples. Then call POST /training/export to create NeMo JSONL.
User: "Sync new traces" → Call POST /traces/sync. Report how many new traces were imported.