Use for R&D program money: multi-period budgets with indirect cost split, burn and runway against milestones, or capitalize vs expense routing. Runs program_budget_planner.py, burn_runway_tracker.py and capex_vs_opex_router.py.

Use when the user asks to deeply read a book, article, PDF, or document set; extract claims and evidence; build a knowledge map; or learn through Feynman explanation and recall. Covers quick, deep, map, Feynman, and whole-book reading modes.

Use for diligence or meeting prep on a specific company, person or organization ('prep me for a meeting with X', 'due diligence on Y'). Tests the user's stated hypothesis with balanced evidence (disconfirming_evidence_balance.py) and writes a .docx dossier.

Use when designing a prospective clinical study: choosing primary and secondary endpoints, estimating sample size and power, or scoring feasibility before submission. Runs endpoint_selector.py, sample_size_estimator.py and phase_gate_scorer.py; outputs are estimates for a biostatistician to confirm.

Use when the user wants a /goal prompt rather than the build itself ('/fable-goal', 'turn this into a goal prompt', 'write the prompt that builds X'). Turns their ramble into one copy-paste prompt checked by goal_prompt_self_check.py.

Use for sprint and team analytics: velocity forecasts, sprint health, retrospectives, burndown or blockers. Forecasts with velocity_analyzer.py (Monte Carlo), scores with sprint_health_scorer.py and tracks retro actions with retrospective_analyzer.py.

Use when the user wants their Claude agent to self-improve from past usage, asks about a nightly/offline 'sleep' or 'dream' cycle, memory/skill consolidation, or says things like 'make my agent better the more I use it', 'review my past sessions', 'learn my preferences', 'consolidate what you learned', 'run the sleep cycle', or wants to schedule offline self-optimization. Drives the skillopt_sleep engine: harvest past sessions → mine recurring tasks → replay offline → consolidate validated CLAUDE.md and SKILL.md behind a held-out gate.

Use when defining or reviewing SLOs, SLIs, error budgets or burn-rate alerts. Designs SLOs with slo_designer.py, computes budgets and multi-window burn-rate thresholds with error_budget_calculator.py and catches common SLO bugs with slo_review.py.

Use for any YouTube URL, video ID or @handle, or a request for transcripts, video search, channel, playlist or new-upload monitoring. Fetches them with yt-dlp, youtube-transcript-api and public RSS feeds, with no API key; no video or audio downloads.

Use when someone asks why their LinkedIn reach dropped, which posts worked, whether a pattern is real or how to test one. Analyzes posts with post_performance_analyzer.py, mines patterns with pattern_miner.py and plans tests with experiment_planner.py.

Use before adding, rolling out or retiring a feature flag or kill switch, or when flag debt piles up. Plans the rollout with rollout_planner.py, finds stale flags with flag_debt_scanner.py and audits kill switches with kill_switch_audit.py.

Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in docs/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.

Use when the user asks what keeps going wrong, after a session where the agent needed repeated correction, or before hardening the environment, to mine past OpenCode sessions in this repository for recurring corrections.

Monitor Nx Cloud CI pipeline and handle self-healing fixes. USE WHEN user says "monitor ci", "watch ci", "ci monitor", "watch ci for this branch", "track ci", "check ci status", wants to track CI status, or needs help with self-healing CI fixes. Prefer this skill over native CI provider tools (gh, glab, etc.) for CI monitoring — it integrates with Nx Cloud self-healing which those tools cannot access.

Windows ML (WinML, Microsoft.Windows.AI.MachineLearning) guidance in three evidence tiers. Tier 1 is canonical Microsoft Learn documentation. Tier 2 is ONNX Runtime / ONNX Runtime GenAI source and NuGet package facts. Tier 3 is Trackdub's own dated observations. Covers model setup, all Windows ML execution providers (CPU, DirectML, OpenVINO, QNN, VitisAI, MIGraphX, NvTensorRtRtx, experimental WebGPU), deployment, ExecutionProviderCatalog install and registration, OrtEpDevice selection, EP-context compilation and caching, ONNX Runtime GenAI on Windows ML, and diagnostics. Use whenever work touches Windows ML or WinML, ExecutionProviderCatalog, EnsureReadyAsync/TryRegister, GetEpDevices, onnxruntime.dll conflicts, Microsoft.ML.OnnxRuntimeGenAI.WinML, DirectML availability, or a Windows ML / ONNX Runtime upgrade, even when the user only says "DirectML is broken" or "why doesn't the NPU show up".

Develop, build, and debug the LangBot core backend and web frontend. Use when working inside the LangBot repository — backend (Python/Quart, src/langbot/pkg), the Vite/React web UI, HTTP API controllers/services, Alembic migrations, or the MCP server. Covers the dev environment (uv, pnpm), repo layout, the API auth model (user token / API key / global key), adding API endpoints, and the rule that API changes must update the MCP server and skills. Triggers on "langbot backend", "langbot dev", "langbot api", "add langbot endpoint", "langbot migration".

Deploy and configure a LangBot instance — Docker / Docker Compose, Kubernetes, the config.yaml model, the Box sandbox runtime, the plugin runtime, and the global API key. Use when installing, deploying, upgrading, or configuring LangBot in production or self-hosted environments. Triggers on "deploy langbot", "langbot docker", "langbot compose", "langbot kubernetes", "langbot config.yaml", "langbot box runtime", "langbot global api key".

Update an OpenSpec change by revising its existing planning artifacts and keeping them coherent with one another. Use when the user wants to revise a change's plan, fold new decisions into it, or reconcile its artifacts after an edit. Never edits code.

Comet 工作流入口。当用户明确调用 /comet，或明确要求使用 Comet 但未指定 Native/Classic 时使用；按项目配置加载 Native 或 Classic。

Generate synthetic PCB inspection images on Kubernetes with the SMT Defect Image Generation (DIG) DAG in CDF Orchestration. The DAG renders components in simulation, turns the renders into realistic golden images (sim2real), and uses AnomalyGen to add texture defects such as solder bridges, excess solder or missing components. Simulation can also render shift, tombstone and sideflip pose defects. Use it to build DIG payloads, pick a pipeline mode (full, simulation, sim2real or defect-only), check GPU readiness, follow a run, and match each defect image to its golden and simulated source. Not for running Simulation, Augmentation or AnomalyGen on their own.