Agent skills you can check before you install
Public SKILL.md skills from GitHub repositories with 50+ stars and from the vendors themselves, scanned and scored. Install any of them with one command in Claude Code, Cursor, Codex, GitHub Copilot or Gemini CLI.
47,891 skills from 2,309 repositories · updated 3 minutes ago
47,891 skills
- deepchat-settingsThinkInAIXYZ/deepchatDeepChat app settings modification (DeepChat 设置/偏好) skill. Activate ONLY when the user explicitly asks to change DeepChat's own settings/preferences (e.g., theme, language, font size...). Do NOT activate for OS/system settings, editor settings, or other apps.WritingNo warnings★ 6.4K↓ —
- computer-useThinkInAIXYZ/deepchatDrive native desktop apps through DeepChat's built-in Computer Use tools. Use when the user asks to operate, inspect, automate, or perform a GUI task in a real desktop application.ProductivityNo warnings★ 6.4K↓ —
- hunt-idorelementalsouls/Claude-BugHunterHunting skill for idor vulnerabilities. Built from 26 public bug bounty reports. Use when hunting idor on any target.SecurityNo warnings★ 5.4K↓ —
- hunt-csrfelementalsouls/Claude-BugHunterHunting skill for csrf vulnerabilities. Built from 15 public bug bounty reports including modern variants — SameSite=Lax sibling-subdomain bypass (Argo CD CVE-2024-22424), GraphQL mutations-via-GET (GitLab $3,370), framework-wide CSRF middleware disabled (Stripe Dashboard $5,000), path-traversal CSRF-token bypass (GitHub Enterprise CVE-2022-23732 $10k), Origin-omission bypass (TikTok $2,500), OAuth-state null-byte (Streamlabs), WebSocket CSRF / CSWSH (Coda), default-SameSite email-change → ATO (YoYo Games $400), social-account-link CSRF (HackerOne), JSON-CSRF via text/plain on email-change (TikTok $500). Use when hunting modern CSRF — heavy emphasis on chain-to-ATO patterns.SecurityNo warnings★ 5.4K↓ —
- ligandmpnnaipoch/open-scienceInverse-fold a backbone with ligand, nucleic-acid, and metal context using LigandMPNN (Dauparas et al. 2023, github.com/dauparas/LigandMPNN). Reach for this skill to redesign the residues lining a binding pocket around a bound small molecule or cofactor, to design metal-coordinating sites where the geometry must be respected, or to get threaded designed-sequence PDBs out of any MPNN run.DesignNo warnings★ 5.4K↓ —
- group-by-analysisOpenSenseNova/SenseNova-Skills对多 Sheet 的 Excel 文件进行行数统计、大文件 Parquet 转换预处理、数据清洗及分组聚合分析，并生成带样式标记的统计表与可视化图表。DocumentsNo warnings★ 6K↓ —
- hunt-sharepointelementalsouls/Claude-BugHunterHunt Microsoft SharePoint Server (2013/2016/2019/Subscription Edition) on-prem farms — anonymous endpoint enumeration, version disclosure, legacy SOAP login bypass (Authentication.asmx), ToolShell precondition chain (CVE-2025-53770), SafeControl reflection enumeration via Picker.aspx, NTLM Type-2 AD topology disclosure, custom-branding module discovery, EoL farm permanent-CVE-window exploitation, FormDigest anonymous issuance, file-extension blocklist NOT-an-oracle pattern, custom-zone Forms auth bridging on-prem AD. Use when target has SharePoint headers (SPRequestGuid, X-MS-InvokeApp, X-SharePointHealthScore, MicrosoftSharePointTeamServices) or paths (/_layouts/15/, /_vti_bin/, /_api/, /_catalogs/).SecurityNo warnings★ 5.4K↓ —
- hunt-shadow-apielementalsouls/Claude-BugHunterHunt shadow / zombie / undocumented API surface (OWASP API9 Improper Inventory Management) — enumerate the full API version history (v1/v2/beta/legacy paths, header- and subdomain-based versioning), pull and diff every reachable OpenAPI/Swagger spec (including ones only findable via the Wayback Machine), and behaviorally diff old vs. current versions for auth/rate-limit/validation regressions rather than just response-shape differences. Distinct from hunt-api-misconfig, which owns exploitation once you have a spec or endpoint (mass assignment, JWT, OData, Swagger-chain attacks); distinct from hunt-subdomain, which owns host-level discovery. This skill owns the version-inventory and behavioral-diff workflow itself. Use when the target has versioned API paths, multiple specs, a changelog referencing deprecated endpoints, or a mobile app whose hardcoded backend calls look older than the current web app's.EngineeringNo warnings★ 5.4K↓ —
- hunt-rag-vectorelementalsouls/Claude-BugHunterHunt vector-store / embedding-layer weaknesses in RAG pipelines (OWASP LLM08 Vector and Embedding Weaknesses) — persistent corpus poisoning that survives across sessions and users (distinct from one-shot indirect prompt injection, which is owned by hunt-llm-ai), cross-tenant vector-database IDOR (unauthenticated or unscoped queries against Pinecone/Weaviate/Chroma/Milvus/Qdrant/pgvector), source-text/metadata leakage in similarity-search results, and retrieval-hijack via adversarial embedding proximity ('SEO poisoning' for RAG). Targets: any app with a shared knowledge base, document upload feeding a chatbot, or a directly reachable vector-DB port. Validate: a second, clean session/account must inherit a poisoned result, or a cross-tenant artifact must be independently verifiable — confabulation is not a finding, same bar as hunt-llm-ai. Use when target is RAG-backed, exposes a vector-DB port, or lets users upload documents that other users' queries later retrieve.SecurityNo warnings★ 5.4K↓ —
- hunt-grpcelementalsouls/Claude-BugHunterHunt gRPC vulnerabilities — server reflection enabled (enumerate all services/methods), missing authentication / metadata-stripping on internal endpoints, plaintext gRPC over HTTP/2, internal endpoint disclosure, proto file leakage, gRPC-Web/grpc-gateway transcoding injection, and HTTP/2 Rapid Reset DoS (CVE-2023-44487). Use when target exposes port 50051 / 443 / 8443 / 9090 with HTTP/2, when grpcurl/grpcui detects reflection, when an Envoy or grpc-gateway proxy is fronting a microservice, or when recon reveals a microservice architecture.SecurityNo warnings★ 5.4K↓ —
- web-artifacts-builderanthropics/skillsSuite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.OfficialDesignNo warnings★ 185K↓ —Also in 2 other repos
- claude-apianthropics/skillsReference for the Claude API / Anthropic SDK — model ids, pricing, params, streaming, tool use, MCP, agents, caching, token counting, model migration. TRIGGER — read BEFORE opening the target file; don't skip because it "looks like a one-liner" — whenever: the prompt names Claude/Anthropic in any form (Claude, Anthropic, Fable, Opus, Sonnet, Haiku, `anthropic`, `@anthropic-ai`, `claude-*`, `us.anthropic.*`, `[1m]`); the user asks about an LLM (pricing/model choice/limits/caching) — never answer from memory; OR the task is LLM-shaped with provider unstated (agent/MCP/tool-definition/multi-agent/RAG/LLM-judge/computer-use; generate/summarize/extract/classify/rewrite/converse over NL; debugging refusals/cutoffs/streaming/tool-calls/tokens). SKIP only when another provider is being worked on (overrides all triggers): OpenAI/GPT/Gemini/Llama/Mistral/Cohere/Ollama named in the query; OR `grep -rE 'openai|langchain_openai|google.generativeai|genai|mistralai|cohere|ollama'` over the project hits (run this grep FIRST OfficialEngineeringNo warnings★ 185K↓ —Also in 1 other repo
- formatted-export-with-parquetOpenSenseNova/SenseNova-Skills从多Sheet Excel文件中识别指定条件的记录，并将筛选结果以整行标红格式导出为Excel文件，适用于数据清洗、条件筛选与可视化标记场景。DocumentsNo warnings★ 6K↓ —
- mirrord-header-propagationmetalbear-co/mirrordMakes a codebase propagate W3C baggage (the `baggage` header carrying `mirrord-session=<key>`) end to end, so mirrord HTTP filters, preview environments, and queue splitting can route a whole request chain to one developer's session. Detects existing OpenTelemetry, Datadog, or other tracing that can already propagate baggage and enables it; otherwise adds a small propagation layer. Covers inbound/outbound HTTP and gRPC, publishing to and consuming from Kafka, RabbitMQ, SQS/SNS, GCP Pub/Sub, Azure Service Bus, NATS, Redis Pub/Sub, BullMQ and Temporal, per-message context restore in consumers, and storage hand-offs such as a GCS signed URL upload that triggers Pub/Sub or an S3 presigned upload that triggers SQS. Ends with a report of every flow found and which ones still drop the header. Use when the user asks to propagate headers, baggage, tracestate, the mirrord session key, or trace context across services or queues, or when a mirrord filter matches the first service but not downstream services or queue consDevOpsNo warnings★ 5.4K↓ —
- openrlhf-trainingOrchestra-Research/AI-Research-SKILLsHigh-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.PlanningNo warnings★ 14.1K↓ —
- mirrord-db-branchingmetalbear-co/mirrordHelps users configure mirrord.json for database branching, enabling isolated database copies for safe development and testing. Use when the user wants to set up MySQL, MariaDB, PostgreSQL, MSSQL, MongoDB, Redis, DynamoDB, ClickHouse, Google Spanner, Amazon S3, turbopuffer, or generic branches, configure copy modes, connection sources, schema migrations, IAM authentication, or manage database branches.TestingNo warnings★ 5.4K↓ —
- replica-reconJakeschincariol/replica-skillReverse-engineers any app into a recon map: screen inventory, user flows, component list, inferred data model and a feature matrix, from public pages, screenshots, app store listings, help docs and the user's own account. The first step of the Replica pack. Use when the user says "clone this app", "reverse engineer X", "how does X work", "map out X", "what screens does X have", "I want to build my own version of X", "copy this app", or pastes an app's URL or App Store link and wants to rebuild it.OtherNo warnings★ 6.6K↓ —
- replica-architectJakeschincariol/replica-skillPlans the stack, database schema and API for an app clone, from the recon map replica-recon wrote. Picks boring, managed tech, turns the inferred data model into real SQL with indexes and access rules, lists every route by flow, and orders the build as a thin vertical slice first. Use when the user says "plan the clone", "what stack should I use", "design the database", "write the schema", "plan the API", "architecture for my version of X", or after /replica-recon finishes.PlanningNo warnings★ 6.6K↓ —
- glaminternet-court/internet-court-skillSolana vault management via GLAM Protocol. Triggers: glam, glam-cli, glam-sdk, vault create/manage, tokenized vault, share class, DeFi vault, treasury, asset management, access control, delegate permissions, Jupiter swap, Kamino lending/borrow/vaults/farms, staking (Marinade/native/SPL/Sanctum/LST), cross-chain USDC (CCTP), timelock, subscription/redemption, NAV pricing, token transfer. Supports CLI and TypeScript SDK.EngineeringNo warnings★ 7K↓ —Also in 1 other repo
- quicknode-skillinternet-court/internet-court-skillQuicknode blockchain infrastructure for 80+ chains: Core RPC API, Streams, Webhooks, Blazar WSS, Solana gRPC, HyperCore gRPC, SQL Explorer, Metaplex DAS API, Blockbook, Ordinals & Runes API, Swap API, x402, MPP, Agent Subscriptions, CLI, and MCP. Use for any Quicknode product, blockchain data access, or wallet-paid agent access.DevOpsNo warnings★ 7K↓ —
- exp-mock-usage-analysisdotnet/skillsAudits .NET test mock usage by tracing each mock setup through the production code's execution path to find dead, unreachable, redundant, or replaceable mocks. Use when the user asks to audit mock usage, find unused or unnecessary mock setups, check if mocks are needed, reduce mock duplication or over-mocking, simplify test setup, or review whether mock configurations like ILogger/IOptions should use real implementations instead. Supports Moq, NSubstitute, and FakeItEasy.TestingNo warnings★ 5.8K↓ —
- melonjs-loading-assetsmelonjs/melonJSUse this skill for loading and managing game assets in melonJS — the resource descriptor format, the full asset type list, preload versus load, retrieving loaded assets, base URLs and cross-origin settings, and the loading screen. Triggers on: loader, loader.preload, loader.load, resources, getImage, getJSON, getTMX, getGLTF, getVideo, getFont, setBaseURL, setOptions, crossOrigin, withCredentials, credentials, cookies, authenticated assets, CORS, asset, preload, LOADER_COMPLETE, loading screen, unload, setParser, custom asset type, generated texture, procedural texture.OtherNo warnings★ 6.4K↓ —
- integration-testsinternet-court/internet-court-skillWrite and run integration tests against a GenLayer environment.TestingNo warnings★ 7K↓ —
- mirrord-temporalmetalbear-co/mirrordHelps DevOps engineers configure mirrord Operator's Temporal task queue splitting feature end-to-end. Generates the MirrordSplitConfig and MirrordPropertyList Kubernetes CRD YAMLs, the matching mirrord.json split_queues section with message_filter and jq_filter, and Helm value guidance. Use this skill whenever the user mentions Temporal splitting with mirrord, Temporal task queue splitting, splitting a Temporal worker, configuring mirrord with Temporal or Temporal Cloud, routing Temporal workflow or activity tasks to a local worker, or troubleshooting Temporal splitting sessions. Also trigger on split_queues with queue_type Temporal, or connecting mirrord to a Temporal frontend. This is an alpha Team/Enterprise feature of mirrord.DevOpsNo warnings★ 5.4K↓ —
openrlhf-training
Orchestra-Research/AI-Research-SKILLs
PlanningUpdated Oct 11, 2026
From the author
High-performance RLHF framework with Ray+vLLM acceleration. Use for PPO, GRPO, RLOO, DPO training of large models (7B-70B+). Built on Ray, vLLM, ZeRO-3. 2× faster than DeepSpeedChat with distributed architecture and GPU resource sharing.
Install
npx skills add Orchestra-Research/AI-Research-SKILLs --skill openrlhf-training -a claude-code
Installs to .claude/skills/openrlhf-training/ · Each command fetches the skill straight from GitHub with the open skills CLI.
Download .zip
Byte-identical to the scanned commit, with the repository’s licence and a SOURCE.txt.
What it can touch
Instructions onlyUses the networkLicence: MIT
Safety checks on SKILL.md
- Uses sudoLow
Why 91
Security95
Provenance90
Maintenance100
Licence100
Adoption65
Files (5)
- SKILL.md
- references/algorithm-comparison.md
- references/custom-rewards.md
- references/hybrid-engine.md
- references/multi-node-training.md