How to launch distributed Megatron-LM training jobs on a SLURM cluster. Covers a minimal sbatch skeleton, environment-variable setup for torch.distributed.run, CUDA_DEVIC — from…
Run a 1:1 that serves the engineer — agenda from them, growth thread, blockers surfaced — not your status update. Reach for this before defaulting to a status read.
Use Qwen Code as a repeatable terminal coding agent: install the CLI, authenticate a provider, run interactive or delegated coding sessions, and connect skills, MCP servers,…
Install, configure, update, health-check, and switch model profiles for OpenCode and the oh-my-openagent (omo) plugin.
Install an OpenCode orchestration plugin that routes codebase work across specialist agents, background tasks, model presets, bundled skills, and live terminal multiplexers.
Build, run, and drive OpenFront locally. Use when asked to run the game, start the dev server, take a screenshot of the UI, verify a client change in the real app, or interact…
Use Claude Code Agent Farm when a repo needs many Claude Code workers to run in parallel on bug-fix or best-practice sweeps with coordination, conflict avoidance, and live…
Use ADHD to fan out isolated reasoning branches for open-ended coding-agent decisions, then score, prune, and deepen the strongest options.
Install a maintained skill pack for weekly planning, retrospectives, task routing, department setup, and founder-ops workflows across Claude Code and Codex.
Use AgentSys to install and run structured coding-agent pipelines for task discovery, delivery preparation, drift checks, review, release, and shipping.
This entry turns WordPress/agent-skills' WP PHPStan guidance into a bounded agent workflow for plugin, theme, and site repositories.
Run and screenshot the portfolio web app. Use when asked to "run the portfolio", "screenshot the home page", or "verify a change in the browser".
PR レビュー全体を1コマンドで実行する skill。スタイル参考ガイド (/pr-review-style-reference) の読み込み・PR 情報取得・レビュー文面作成・post-pr-review での投稿・resolve-pr-threads での過去スレッド整理までを通しで行う。caller (GitHub Actions など) からは本…
Use Text Extract API when an agent needs to turn PDFs, Office files, or images into Markdown or structured JSON with local OCR, optional Ollama models, PII removal, and queued…
Use Hive as a model-agnostic harness for long-running multi-agent workflows that need state, recovery, observability, human oversight, and business-system tools.
Execute named restic backup profiles with repeatable backup, retention, prune, check, and restore steps instead of hand-running one-off commands.
Use FuXi when an operator wants a terminal coding agent to edit code, run commands, drive tools, and verify changes with cost-aware routing across LLM providers.
Create and run an autonomous overnight task queue. Use when the user asks to "run tasks overnight", "queue up work", "create an overnight queue", "batch these tasks", "run this…
Dispatch long-running coding work to background agents, check progress later, and pull reviewed outputs back into the main repo flow instead of babysitting one foreground session.
Ingest a bounded document set into Verba, configure retrieval and model providers, then review grounded chatbot answers before handing context to an agent workflow.
Run benchmark-style eval suites against models or agents, then inspect scored traces instead of relying on ad hoc chats and gut feel.
Run git-slop report, explain, plan, and check commands in a repository while preserving its local artifact and warn-only conventions.
Use Codemod CLI when an agent needs to scaffold, test, and run a reviewable multi-step migration workflow with approval gates, rather than applying a one-off search-and-replace or…
Use when the user asks for a long-running Robot Networks listener. Streams live envelope.notify push frames from the ASMTP network.
Run, smoke-test, and drive the SAP Router Skill CLI (sap_router.py, memory_manager.py, xls_to_bapi.py, template_repo.py, abap_serializer.py, cpi_iflow_packager.py).
Use OpenFang to run packaged autonomous agent workflows on schedules with manifests, dashboards, reporting, and guardrails for sensitive actions.
Launch the Investment Screener suite (backend + frontend). Trigger when user says "run the screener", "start the app", "launch investment toolkit", "start servers", or "run…
Navigate the AI-SDLC workflow for a given work item or artifact: report where it is in the lifecycle and what to do next.
Use IronClaw Reborn as a secure local agent runtime for model-routed CLI, REPL, and WebUI sessions with explicit state, policy, storage, and secret boundaries.
Use Codeman to supervise persistent Claude Code, Codex, OpenCode, and Gemini CLI sessions through a self-hosted web dashboard with live terminals and API controls.
Run-book for executing a Seestar S50 imaging session via the seestar-mcp tools. Use whenever the user wants to start, monitor, or wind down an imaging session — e.g.
Coordinate multiple coding agents in one local-first workspace with shared memory, tools, files, browser access, and MCP-backed integrations.
Execute shell commands for system operations, directory creation, and script execution. When you need to run system commands, execute scripts, install packages, or manage…
Build and launch the app in the iOS Simulator. Automatically selects an appropriate simulator device, boots it if needed, and installs and launches the app.
Use to create or update specification docs in docs/specs/ — checks alignment with parent domain knowledge and design decisions
Run a spec-vs-code audit on this codebase. Compare actual implementation in app/, lib/, lib/typeorm/, typeorm/, and tests against project specifications in docs/.
Use OSpec to turn a coding request into a repo-persisted spec workflow with initialization, change advancement, verification evidence, and archival.
Use MoAI-ADK to install Claude Code agents and skills for spec-first implementation, TDD, DDD, and quality-gated development.
Use Ouroboros when a coding agent should turn a vague request into a replayable spec, execute it across supported runtimes, and pass explicit evaluation gates.
Build, launch, and drive the SredstvaApp WPF desktop app (screenshot, click, type, log in) via a UI Automation PowerShell driver.
Use when the user says "draft my standup" / "what's my standup today" / "Monday standup" — pulls recent commits and PR activity from the connected code host, pulls recently closed…
Use CloakBrowser as a supervised Playwright/Puppeteer-compatible browser runtime for high-friction browser automation and anti-bot test environments.
Use Council of High Intelligence when an agent should convene a structured panel, force disagreement, and return a decision with dissent, kill criteria, and next actions.
Install StyleSeed skills so coding agents choose a design grammar, build UI with project-local rules, score the result, and verify rendered screens before handoff.
Install Suede Creator Skills when an agent needs repeatable Codex, Claude Code, or MCP-backed workflows for orchestration, code review, ship gates, AI evals, design, copy, SEO,…
Use Warp as a supervised terminal workspace for running Warp Agent or external CLI agents such as Codex, Claude Code, Gemini CLI, and OpenCode with shared shell context.
Give an MCP-capable agent a local Playwright browser with human takeover, reusable auth profiles, approvals, audit trails, and repeatable browser workflow evidence.
Start a bounded Suna or Kortix agent session, let it use browser, files, tools, and connectors, then review the session output before delivery.
Use Empryo when a coding agent should map symbols, blast radius, and dependencies before making AST-backed edits across a repository.
Run or rebuild the local Tamagui/Next.js frontend (frontends/tamagui). Use when the user wants to start the Tamagui dev server or rebuild it standalone.
Use GitHub Copilot CLI to plan, edit, debug, explain, and automate repository work from a terminal with GitHub context and approval controls.
Lets an agent use Gemini from the terminal to inspect repositories, edit files, run shell commands, and ground work with built-in search and local context.
Run unit and integration tests for Catalyst-Relay. Use when asked to test, run tests, verify changes, or check if code works.
Use at the end of an evaluation — would this pitch still be on your mind at dinner, or does it blur into the other AI/SaaS pitches? Name the one compounding idea, and the single…
Use the moment prod is broken/degraded or an alert fires — declare an incident, take command, mitigate before diagnosing, and communicate on a cadence.
Turn a planned backlog into markdown tickets, dispatch them through coding agents, and only get pulled back in when a blocker or decision really needs you.
Run Agentic Tree Search on a coding or design task with explicit branch tracking and verifier-gated finalization.
Submit ML training to Databricks and pull results. Use when the user wants to run, train, or execute a model on Databricks.
Use the Fable Method skill bundle to classify work, gather evidence, act narrowly, verify by observation, and judge agent outputs against repeatable eval-backed rules.
Build and launch the TianWen.Cli TUI in a new Windows console window with stderr redirected to tui-stderr.log.