Намерете инструкция, прочетете оригиналния файл и подгответе команда за своя инструмент.
31 301 skills в каталога
Как подбирам и измервамДанните имат нужда от нова проверка.Последна успешна проверка: 23.09.2026 г.
Проверката на източника не означава, че всяка инструкция е изпробвана от Level8. Каталогът не обхваща непременно всички публични хранилища.
Търсене в целия индекс. Метаданните показват посоченото от източника; липсващите стойности остават означени.
Датата в GitHub е отделна от първата поява в моя индекс. Филтърът по дата изключва записите без посочено обновяване.
Страница 148: 24 записа в 16 групи по репозиторий. Всички записи от страницата са показани. Номерът пази позицията от избраната подредба.
1 skill от този източник на тази страница
自动从邮箱收集报销发票并汇总到 Excel。扫描 IMAP 邮箱指定文件夹，下载 PDF 附件与正文链接里的电子发票（含诺诺/百望 SPA、51发票中间页、税局直下链接），解析发票号/金额/购销方/日期等字段，按发票号去重追加进一张汇总表。当用户提到报销、发票、邮箱收发票、发票下载、发票整理、invoice、fapiao、把发票汇总到 Excel 时使用。
1 skill от този източник на тази страница
Use when an agent needs to understand how an unfamiliar codebase or subsystem actually works before changing it; maps entry points, control and data flow, key abstractions, and external effects to the depth the task needs, and reports what was not traced rather than reviewing quality or editing code.
1 skill от този източник на тази страница
Fetch and map eBay and Facebook Marketplace listings into Deal Sniper Listing models via eBay API and Apify. Use when working on ingestion, scrapers, ebay.py, apify.py, or marketplace adapters.
7 skills от този източник на тази страница
Develop Fabric mods for Minecraft 1.20.1, 1.21.x and 26.x. Project setup with the fabric-loom Gradle plugin (loom 1.17.19, loader 0.19.3, Fabric API per version), fabric.mod.json entrypoints and depends blocks, official Mojang mappings, Mixin, client/common source sets, and the build/runClient/runServer loop. Use when creating or extending a Fabric mod, or fixing Fabric build, remap or dependency problems.
Develop Forge mods across the four eras: 1.7.10 (ForgeGradle 1.2 fork, Gradle 7.4.2, MCP), 1.12.2 (ForgeGradle 3, Gradle 4.9, MCP snapshot 20171003), 1.16.5 (ForgeGradle 5, Gradle 7.3.3, official mappings), 1.20.1 (ForgeGradle 6, Gradle 8.8, mods.toml). Covers mcmod.info vs mods.toml metadata, @Mod lifecycle per era, reobfJar, JDK 8 requirement for old lines, runClient/runServer/runData. Use when creating or extending a Forge mod, or fixing era-specific Forge build problems.
Reference for addon development against major mods on the legacy lines: Thaumcraft, Tinkers Construct, Botania, Twilight Forest, Applied Energistics 2, Mekanism, IndustrialCraft 2, Thermal Expansion, Draconic Evolution and SlashBlade for Minecraft 1.7.10 and 1.12.2. Covers the general addon patterns: soft-dependency detection across Forge/Cauldron/Spigot, compileOnly optional dependencies, curse.maven and Modrinth maven coordinates per mod, and recipe/event integration hooks. Per-mod coordinates
Develop NeoForge mods for 1.20.1, 1.21.x and 26.2. Project setup with the net.neoforged.moddev Gradle plugin (2.0.144; legacyforge 2.0.91 for 1.20.1), NeoForm pipeline, neoforge.mods.toml mandatory fields, @Mod and event bus, DeferredRegister, data generation, runClient/runServer/runData, optional Parchment mappings. Use when creating or extending a NeoForge mod, or fixing moddev/NeoForm/mods.toml problems.
Develop Spigot/Bukkit server plugins for legacy Minecraft 1.7.10 and 1.12.2: Java 8 build setup with spigot-api compileOnly (1.12.2 from the Spigot hub maven, 1.7.10 from a vendored server jar), plugin.yml fields without api-version, the JavaPlugin lifecycle, events, commands, permissions, config, and the build/copy-to-server/reload loop on old servers. Also covers Cauldron/KCauldron/Thermos hybrid servers (Forge mods + Bukkit plugins on one 1.7.10 server). Use when creating or extending a plugi
Four-stage Minecraft delivery team for Codex: A plans and web-verifies, B scaffolds the project with mc_scaffold, C fills feature content, D builds and reviews with mc_gradle. The stages run as fresh subagents in strict A -> B -> C -> D order and share exactly one artifact, <project>/PLAN.md. Use when the user asks for a whole plugin/mod/addon with 3 or more work items, or explicitly asks for the staged/team workflow; small edits should stay inline.
Four-stage Minecraft delivery team for Codex: A plans and web-verifies, B scaffolds the project with mc_scaffold, C fills feature content, D builds and reviews with mc_gradle. The stages run as fresh subagents in strict A -> B -> C -> D order and share exactly one artifact, <project>/PLAN.md. Use when the user asks for a whole plugin/mod/addon with 3 or more work items, or explicitly asks for the staged/team workflow; small edits should stay inline.
2 skills от този източник на тази страница
Generate Minecraft-item-style pixel art textures via an OpenAI-compatible image API. Reads IMAGE_BASE_URL, IMAGE_MODEL, and IMAGE_API_KEY from a .env file, supports text-to-image and image-to-image (with one or more reference images), constrains the prompt with a pixel-art preamble, then downsamples the output to caller-specified target sizes (16x16, 32x32, 64x64, etc.) with optional palette quantisation. Use this skill whenever the user asks for a Minecraft item/block/tool icon, mod texture, sp
Find and install Minecraft mods as compile/runtime dependencies in a NeoForge/Forge Gradle project. Use this skill when the user asks to add a mod dependency, integrate with another mod, add mod compat, or reference another mod's API. Trigger on phrases like "add X dependency", "integrate with X mod", "add X as a library", "add compat for X", or when the user names a specific Minecraft mod and wants to use its API.
1 skill от този източник на тази страница
General workflow for auditing ML CI failures, experiment regressions, training run failures, golden metric failures, and telemetry-backed ML work-product claims from local repositories, logs, metrics, configs, and artifacts. Use when Codex needs to decide whether an ML failure is a model/convergence issue, correctness bug, data/config issue, infrastructure/runtime issue, evaluation/gating policy issue, or unsupported claim, and produce structured evidence-backed outputs.
2 skills от този източник на тази страница
Keyless agent identity and trust score, no account needed. Register a DID with an Ed25519 keypair and proof of work, bind an API key, issue a verifiable credential, verify any agent and read its reputation.
Keyless agent identity and trust score, no account needed. Register a DID with an Ed25519 keypair and proof of work, bind an API key, issue a verifiable credential, verify any agent and read its reputation.
1 skill от този източник на тази страница
Orchestrate parallel browser automation using multiple Chrome DevTools MCP instances. Use when a task requires doing the same browser action across many targets simultaneously (e.g., submitting contact forms, filling applications, scraping pages that need JS rendering). Spin up 1-5 parallel Chrome agents.
1 skill от този източник на тази страница
This skill should be used when a coding job splits into several independent, mechanical edits that can be delegated to Muse Code (`muse exec`) workers in isolated git worktrees, each supervised by an Opus agent that runs the acceptance check and re-prompts until the patch is right. Use it for repeti
1 skill от този източник на тази страница
Carry herdr worktree spaces and their Claude sessions from this machine to another host (e.g. user@far-host), open them as herdr spaces there, and PROVE the crossing with the 4 locks — interactively, the human picks from a list. Use when the user says 'noah bring', 'maw noah bring', 'bring X up to <host>', 'ferry these worktrees to <host>', 'ขน worktree ไป <host>', or names herdr spaces plus a host. Same-ghq-root hops only (both hosts share one ghq root, e.g. /opt/Code). Do NOT use when the ghq
1 skill от този източник на тази страница
Node.js development environment. Use when the project needs this capability or the user / team manifest asks for it. Use for specialized `nodejs` work when listed in TEAM.yaml or explicitly requested.
1 skill от този източник на тази страница
Instruments code so production behavior is visible and diagnosable. Use when adding logging, metrics, tracing, or alerting. Use when shipping any feature that runs in production and you need evidence it works. Use when production issues are reported but you can't tell what happened from the available data.
1 skill от този източник на тази страница
Use when archiving bare URLs or Obsidian: payloads. Delegate links, files, text and code to the library.
1 skill от този източник на тази страница
Build or debug an app that has to keep working with no network — local writes, a queue, and reconciliation on reconnect. Use when adding offline support, when work is lost on a dropped connection, when sync produces duplicates or silently drops records, or when photos and other large files need to survive being captured out of coverage.
1 skill от този източник на тази страница
Generates a project onboarding website by deep-analyzing the current repository — no interview, no questions. Reads code, config, routes, data models and docs to derive a single self-contained onboarding.html with fixed sections — title, description, main goal, problem it solves, target user types, high-level system flowchart, sitemap, ERD / schema, tech stack. Stack-agnostic (works on any language/framework). Diagrams render with Mermaid.js via CDN. Output lands at repo root and is added to .gi
1 skill от този източник на тази страница
Harden a plan with an independent OpenCode review, then build it and cross-inspect the result. Claude Code coordinates recon, requirements and PLAN.md; OpenCode reviews the plan in a read-only session; whoever builds never inspects. Use for openclaude this feature, openclaude this plan, review this plan with opencode, build this with opencode and have claude review it, or /openclaude-loop. Not for trivial edits.
Пазя съществуващия каталог и добавям намерени SKILL.md файлове след прочит на метаданните. Наличният индекс остава използваем, когато проверката е частична.
Платформата и нивото описват записаното в източника. Те не са гаранция за съвместимост. Български текст се показва само когато в каталога има отделен превод; иначе виждате оригинала с означен език.
Общите звезди показват натрупана популярност. За „Набират интерес“ се сравнява измереният темп при достатъчен интервал. ≥ означава непълна извадка. Под един час не изчислявам дневен темп; методът и изминалият период се показват до записа.
Проверете файла, лиценза и нужния достъп преди употреба. Нито каталогът, нито четенето на SKILL.md изпълняват външни инструкции.
Темата Agent Skills в GitHub