AI & ML

langwatch Skill

Test your AI agent with simulation-based scenarios. Covers writing scenario test code (Scenario SDK), creating platform scenarios via the `langwatch` CLI against a connected agent, reading the run parameters that agent declares so the scenarios and comparison runs turn its real levers, and red teaming for security vulnerabilities. Auto-detects whether to use code or platform approach based on context.

langwatch Skill

Create and run LangWatch experiments for pre-deployment batch testing. Use when the user wants to test an agent against a dataset, compare prompts or models, benchmark quality, detect regressions, or add a CI quality gate. Do not use for production monitoring or guardrails.

langwatch Skill

Connect the codebase's AI agent to LangWatch agent simulations, so test suites run against the real agent process. Adds a small connect function beside the service startup that calls the agent already in the codebase, which opens an outbound connection and registers the agent with its environment and its run parameters, confirms the agent is Online, and runs the first test suite. Falls back to an HTTP registration when the agent cannot import the SDK. Use when the user wants platform scenarios to test their real agent.

langwatch Skill

Investigates the context economics of your own coding-agent sessions in LangWatch. Reads real sessions to find where carrying a fat context stops paying for itself, measured in cache rebuilds, compactions and cost per turn, and delivers a report with the context size your sessions should stay under, with example sessions behind every claim. Use when coding-agent sessions feel expensive or degrade as they grow.

langwatch Skill

Prices your real LangWatch usage mix against other model providers. Exports your actual token mix per model, including the cache read and write split, fetches current price cards, and reprices the same month of usage under each candidate, with the cache sensitivity stated. Use when someone asks whether a cheaper provider or model would actually save money on your workload.

langwatch Skill

Root-cause production errors and misbehaving agent runs with LangWatch. Finds errored traces, inspects spans, checks monitor and evaluator scores, then narrows to a root cause. Use when something is failing or misbehaving in production (errors, bad answers, latency spikes).

kyopark2014 Bundle

Desktop GUI control via screenshot + mouse/keyboard (Computer Use). Use when the user asks to operate native apps, click UI outside the browser, automate desktop workflows, take/analyze the screen, or mentions computer use / desktop automation / GUI agent / 화면 조작 / 데스크톱 자동화. Prefer browser-use for web-only tasks.

kyopark2014 Bundle

Control macOS via AppleScript / JXA using the mac-operator skill CLI (and optional MCP). Use when the user asks to operate Finder, Safari, Notes, Freeform/Whiteboard, clipboard, notifications, volume, Spotlight, menus, keystrokes, open/quit apps, or any native Mac automation (mac 조작, AppleScript, 자동화, 파인더, 화이트보드, 클립보드, 알림). Prefer browser-use for heavy web DOM work and computer-use for pixel drawing / screenshot loops when scripting dictionaries are insufficient. For "앱 열고 도형 그리기" use mac-operator to open the app, then computer-use to draw.

materialofair Bundle

Test a OpenCode skill with prompt scenarios, invocation checks, negative cases, and execution validation.

materialofair Bundle

Analyze raw prompts, identify intent and gaps, match ECC components (skills/commands/agents/hooks), and output a ready-to-paste optimized prompt. Advisory role only — never executes the task itself. TRIGGER when: user says "optimize prompt", "improve my prompt", "how to write a prompt for", "help me prompt", "rewrite this prompt", or explicitly asks to enhance prompt quality. Also triggers on Chinese equivalents: "优化prompt", "改进prompt", "怎么写prompt", "帮我优化这个指令". Specially handles short Chinese bug reports (the dominant real-world use case): runs Bug Report Triage to extract repro/expected/actual/environment, asks up to 3 clarifying questions if ≤ 2 fields are present, and inserts systematic-debugging as a hard prerequisite before any code change. DO NOT TRIGGER when: user wants the task executed directly, or says "just do it" / "直接做". DO NOT TRIGGER when user says "优化代码", "优化性能", "optimize performance", "optimize this code" — those are refactoring/performance tasks, not prompt optimization (unless the user exp

yokingma Skill

Use when the user needs browser automation for websites, web apps, screenshots, form filling, scraping, QA, exploratory testing, Electron apps, or Slack automation.

mrilikecoding Bundle

Workflow architect that decomposes complex tasks, maximizing local Ollama model leverage via llm-orc while maintaining quality through reflective evaluation.

mrilikecoding Skill

Ensemble architect that composes purpose-built DAGs of local models and scripts, accumulating design knowledge through systematic experimentation via llm-orc.

surya8991 Skill

Packs the entire codebase (or selected folders) into a single token-efficient file using repomix, so other skills can analyze the whole repo without re-reading individual files. Use when the task requires whole-codebase context: full security audits, cross-cutting refactors, architecture review, project-wide code review, or first-time onboarding to an unfamiliar repo. Auto-runs once per session at first invocation of agent-master.

amaledtuning Bundle

Design onboarding, first-run experiences, walkthroughs, empty states, product explainers, training flows, explainer decks, educational PDFs, and feature introductions around accurate user mental models. Use when Codex needs to help new users understand what something is, how it works, why it matters, what to do first, or how to update an incorrect expectation.

amaledtuning Skill

Use when the user asks to move to a new chat, prepare a checkpoint, preserve current chat state, or generate a continuation prompt. Applies across projects and should capture current work, files to read next, blockers, validation state, active skills, and next steps without leaking secrets.

dor-rondel Skill

Use this skill when launching the local LangGraph visual development server (Studio), verifying LangSmith tracing environment variables, or testing agent telemetry.

dor-rondel Skill

Use this skill when managing a Python LangGraph repository using uv and a Makefile. Covers dependency syncing, running quality controls (lint, format, typecheck, tests), generating agent graphs, and executing the git commit/push workflow.

fal3 Bundle

Design, implement, diagnose, and test production schema migrations for SwiftData and Core Data, including VersionedSchema snapshots, SchemaMigrationPlan stages, lightweight and custom Core Data mappings, gradual adoption, CloudKit constraints, and forward-recovery planning. Use when a persistent model changes, an existing store must upgrade without data loss, a team is choosing between SwiftData and Core Data for migration requirements, or copied production stores need validation. Do not use for ordinary fetch/query code, transient in-memory model transformations, unrelated file-format migrations, or advice that assumes Core Data should be replaced merely because SwiftData exists.

johnsonbuilds Skill

CLI tool to find which LLM models run on your local hardware — checks GPU/CPU/RAM and returns compatible models

johnsonbuilds Skill

261 hand-drawn style prompt catalog with numbered system — supports GPT Image, Midjourney, and other AI image generators

johnsonbuilds Skill

Compress AI Agent tool outputs, logs, and RAG content using Headroom — reduce token usage by 20-95%

johnsonbuilds Bundle

Scan newly created or modified Hermes skills with NVIDIA SkillSpector before committing — detect prompt injection, data exfiltration, and other security issues. Skips if SkillSpector is not installed.

johnsonbuilds Bundle

Worklittle Jobs MCP — search 4M+ jobs with visa/salary/distance filters, swipe to apply, create resumes & cover letters, and connect your Worklittle account

jpvarbed Bundle

Build or materially redesign an agent skill from a rough idea or a workflow that just succeeded, with a spec-first design, risk-based evidence, real baseline and with-skill evaluation, blind cross-family forward tests, and a typed completion receipt. Use for “create/add a skill,” “turn what we just did into a skill,” or “redesign a skill's behavior, setup, scripts, or eval.” NOT for prose-only edits (use writing-great-skills), static scoring (use linting-and-scoring), prompt-to-script audits (use determinize-refactor), or installing an existing skill (use skill-installer).

jpvarbed Bundle

Analyze a prompt-heavy skill/plugin and produce a prioritized migration plan that moves deterministic instructions out of prose into scripts or structured contracts — improving reliability and cutting token cost. Use when the user says "reduce prompt tokens", "make this skill deterministic", "script-mode this skill", or "audit a skill for token bloat". Outputs a Markdown audit report with a token summary and per-file conversion plan, NOT code. NOT for authoring a skill from scratch (use writing-great-skills) or scoring quality (use linting-and-scoring).

jpvarbed Bundle

Audit the layered instruction stack (in-conversation user → soul.md/global → project guide → skill → tool/system) for conflicting or ambiguous directives, and surface which layer should win. Use when the user says "check for instruction conflicts", "do my layers contradict", "audit soul.md vs project/skill", "why is the agent ignoring X", or before relying on a deep skill stack. NOT for scoring one skill's quality (use linting-and-scoring) or auditing whether a skill obeys itself (use the adherence audit). Built from ManyIH (arXiv:2604.09443) via apply-paper.

meshmakers Bundle

Develop and extend the OctoMesh MCP server (octo-mcp-service) — the Model Context Protocol server exposing ~181 tools that mirror octo-cli plus generic CK CRUD and aggregation/stream-data queries, used by AI assistants to administer OctoMesh tenants without the CLI or GraphQL. Use when adding or modifying MCP tools, classifying tool risk, wiring *ClientContext helpers, building file-transfer flows, or following the mandatory test conventions. Trigger on: MCP tool development, octo-mcp-service, adding MCP tools, McpRisk, MCP server, model context protocol server work in OctoMesh.

meshmakers Bundle

Visualizes OctoMesh Construction Kits as UML class diagrams on a Miro board. Reads CK YAML definitions from a specified git branch in the local octo-construction-kit checkout (non-destructive via git show) and creates one detailed UML diagram per CK — types with attributes and datatypes, records, enums, inheritance arrows, associations with multiplicity, and cross-CK references. Lets the user pick a branch (default main) and choose between a new Miro board or adding to an existing board URL. Trigger on - visualize CK, visualize construction kit, CK diagram, CK to Miro, Miro board for CK, UML diagram for construction kit, CK model visualization, octo construction kit visualization, draw CK, show CK on Miro, CK class diagram.

muhammadali-k Bundle

Run an explicit Codex boss → worker → independent reviewer build loop. Use only when the user invokes `$route`, types `/route`, says “route this,” or explicitly requests the routed multi-agent loop. The root Codex agent owns the plan, permissions, integration, and final verdict; one worker implements in the authorized workspace; a fresh reviewer inspects the result against the plan; the worker fixes validated findings until the root approves or reports a bounded non-convergence. Works for code, analysis, extraction pipelines, and generated files. Do not invoke implicitly for ordinary tasks.

muhammadali-k Bundle

Peer-review a research abstract or manuscript in a Word (.docx) file and return real tracked changes, threaded comments, and a concise reviewer summary. Use for a first review or a re-review against prior comment threads; load the relevant profile for evidence syntheses, trials, observational studies, other clinical designs, or AI/ML/LLM work. Trigger on requests to review, mark up, track changes, assess a revision, or check a .docx against a named venue or reporting standard. Use this when the deliverable is a marked-up .docx; use manuscript-srma for drafting SR/MA/NMA prose without Word review markup. Do not use for PDF-only markup because tracked changes and threaded Word comments require DOCX.

muhammadali-k Bundle

Draft, organize, or revise Muhammad Ali Khan's own research manuscripts and conference abstracts in his academic voice and house organization. Use for clinical/epidemiology cohorts, AI/informatics and LLM-methods papers, and systematic reviews, pairwise meta-analyses, or network meta-analyses when the user asks to draft, continue, restructure, tighten, or make prose read like their prior work. Apply the bundled voice, section, quantitative-reporting, abstract, and skeleton-first/tables-first references. For evidence-synthesis reporting mechanics, compose with manuscript-srma. Do not use to mark up someone else's DOCX with tracked changes and comments; use abstract-review for that deliverable.

prashantdivate Bundle

One-router Yocto Project skill suite for AI coding agents. Use when a request mentions Yocto, OpenEmbedded, Poky, BitBake, embedded Linux builds, BSPs, board flashing, QEMU boot tests, CI/release automation, security hardening, CVE review, secure boot, OTA updates, firewall policy, or when the right specialized Yocto skill is unclear.

scouzi1966 Bundle

Maintain and extend AFM (maclocal-api), a Swift OpenAI-compatible local LLM server and CLI for Apple Foundation Models, MLX models, API gateway proxying, and Vision OCR. Use when working on AFM CLI commands (`afm`, `afm mlx`, `afm vision`), OpenAI `/v1/chat/completions` and `/v1/models` behavior, streaming SSE, tool-calling, structured outputs, reasoning extraction, AFMKit integration, WebUI packaging, or AFM build/test/release scripts.

scouzi1966 Bundle

Run the maclocal-api (AFM/MLX) test suite — automated assertions and smart analysis. Use when asked to test, validate, regression-check, or benchmark AFM before release, after code changes, or for model onboarding.

scouzi1966 Bundle

Add support for a new HuggingFace MLX model to AFM. Use when user wants to add, onboard, or check compatibility of a model — handles everything from "already supported" to implementing new architectures.