Claude Skills by agentskillexchange

Use RagaAI Catalyst to instrument agent runs, collect traces, manage evaluation datasets, inspect execution graphs, and review regressions before deployment.

Use ParseBench to run repeatable document-parser evaluations before an agent relies on PDF, table, chart, or enterprise-document output for downstream decisions.

Use SandboxFusion to run and judge LLM-generated code in controlled sandboxes across many languages and benchmark-style evaluation tasks.

Turn recurring agent lessons, logs, and strategy tweaks into reviewable GEP evolution artifacts that can be applied across supported coding-agent runtimes.

Run CLIProxyAPI when an operator needs a local OpenAI, Gemini, Claude, Codex, or Grok-compatible API facade over authenticated CLI agent access for repeatable agent and tool integrations.

Use DataHub metadata, ownership, schema, and lineage context to ground agent-assisted data discovery and governance workflows.

Use Azure Data API Builder to generate secure REST, GraphQL, and SQL MCP access for supported databases so agents can query and operate data through bounded tools.

Run pREST in front of existing Postgres databases to provide instant REST APIs and a read-only MCP endpoint with auth, ACL, custom SQL routes, and multi-database support.

Use Gortex to index one or more repositories into a local code graph and expose symbol, call-chain, blast-radius, review, and refactoring tools through MCP, CLI, API, or web UI.

Use Zerox to convert PDFs, images, and office documents into Markdown or structured extraction outputs using vision models.

Use LandingAI's ADE CLI to parse visually complex documents, cache parse artifacts, and extract schema-shaped fields with page and bounding-box evidence from an agent or CI workflow.

Use Famulor's OAuth-secured, read-only MCP profile to inspect assistant configurations, versions, workspace catalogs, and unified call, messaging, and email history.

faster-whisper is a reimplementation of OpenAI's Whisper model using CTranslate2 that delivers up to 4x faster transcription with lower memory usage. It supports CPU and GPU inference with 8-bit quantization, batch processing, word-level timestamps, and VAD filtering for accurate speech-to-text conversion.

Use Mydentify's public intent and product feeds to find software by the outcome a person wants, then return canonical evidence-backed matches.

Translate a vague request or GitHub issue into a concrete, testable spec using five structured elements (problem, scope, acceptance criteria, edge cases, success criterion) before writing any code. Prevents scope creep, PR churn, and mid-implementation pivots.

Run a self-hosted terminal agent or custom Pydantic AI harness that can branch a coding or research run, test isolated alternatives, and merge the winning branch.

Safely synchronize free-playable NetEase Private FM tracks into an append-only playlist using a native Rust CLI/TUI. Use when an operator needs QR login, read-only preview, strict playability checks, explicit cross-platform review, append-only sync, or zero-LLM scheduled execution.

Use Superagent to add prompt-injection blocking, PII redaction, repository scanning, and compliance evidence around AI agent runs.

Turn failing AI-agent production traces into hermetic regression cases and CI gates with Tracely.

Use FlagEmbedding to choose BGE embedding or reranking models, encode documents and queries, evaluate retrieval quality, and feed stronger context into RAG workflows.

Use Swag when an agent needs to turn Go handler annotations into reviewable Swagger 2.0 documentation before API docs, clients, or release checks depend on it.

Use a documented design skill to produce HTML-native prototypes, slides, motion demos, visualizations, and review passes from real brand or project context.

Discovers current Atlas Cloud image models, validates model-specific schemas, submits a single asynchronous image-generation request, polls predictions with a finite budget, and verifies downloaded image artifacts. Use when an agent needs text-to-image or image-editing access through the Atlas Cloud API.

Discovers current MuAPI image models, builds a model-specific request, and validates asynchronous image artifacts. Use when an agent needs hosted image generation through one MUAPI_API_KEY, bounded prediction polling, and a credential-free HTTPS download.

Use Auto Subs when an operator or media agent needs local-first subtitle generation, editing, and handoff into DaVinci Resolve, Premiere Pro, After Effects, or a command-line video pipeline.

Use DeepWiki-Open to analyze a GitHub, GitLab, or Bitbucket repository and generate a navigable AI-written wiki with code structure explanations and diagrams for onboarding or agent context.

Use Guizang Social Card Skill from Claude Code or Codex to turn source material into Rednote/Xiaohongshu carousel cards and WeChat cover image pairs.

Use gpt-image2-ppt-skills to have an agent create image-first or editable PowerPoint decks with gpt-image-2, bundled styles, template cloning, and render review.

Use SQLCoder when an analytics agent needs to turn natural-language questions plus schema context into SQL drafts that can be reviewed, validated, and run inside a controlled data workflow.

Install CocoIndex Code as a skill, plugin, CLI, or MCP server so coding agents can initialize, update, and query local AST-backed semantic code indexes.

Use Memanto to add local or cloud-backed persistent memory to Claude Code, Cursor, Codex, and other coding agents through remember, recall, answer, upload, and integration commands.

Initialize and maintain a repo-native Markdown memory layer so coding agents preserve durable decisions, requirements, constraints, and rationale across sessions.

Install CodeSeek to index a repository, expose semantic search and call-graph queries, and register MCP tools for Claude Code or Codex coding agents.

Use SWE-ReX when a coding agent needs a consistent runtime interface for local, containerized, or remote shell sessions with command output, exit codes, interactive tools, and parallel execution.

Use MARM Memory to give Claude Code, Cursor, Codex, Gemini, and other MCP clients a local-first shared memory server with sessions, semantic recall, notebooks, summaries, and code graph context.

Connect RushDB's MCP server so agents can store, search, update, and traverse persistent structured memory without hand-building a separate vector and graph stack.

Expose LSP-backed definition, reference, rename, and diagnostic tools to MCP clients so coding agents can inspect code semantics before editing.

Connect Claude, Codex, Cursor, ChatGPT, or any MCP-capable client to a local-first Markdown knowledge graph so agents can write, search, and reuse project context across sessions.

Add opencode-mem to OpenCode so coding sessions can store, search, and reuse project memories through a local SQLite and vector index.

Use rails-ai-context to expose a running Rails app's schema, routes, models, controllers, conventions, and test patterns through MCP tools so coding agents verify before editing.

Connect coding agents to a shared Stash knowledge base so sessions, files, pages, and team context can be searched, cited, updated, and reused across agent runs.

Install Agent Vision Toolkit so shell-capable coding agents can inspect screenshots, run OCR, locate UI elements, restore interfaces, and operate GUIs through repeatable local vision playbooks.

Drop a `.gaai/` folder into a project so agents separate scope discovery from isolated delivery, with acceptance criteria, memory, QA gates, and daemonized execution.

gptme is an open-source Python framework for running LLM-powered agents in your terminal with access to shell commands, file editing, browser control, Python execution, and persistent memory. It supports OpenAI, Anthropic, Google Gemini, local models via Ollama and llama.cpp, and OpenRouter, making it suitable for offline, privacy-focused, or multi-provider agentic workflows. Built for autonomous operation with session persistence and tool-native agent loops.

Use claudex-loop when a Claude Code operator should lock intent, stress-test a plan with Codex, and cross-inspect the final build before approving risky work.

Use Hermes Tweet when Hermes Agent needs a native X/Twitter plugin for reading posts, exploring public data, and running explicitly gated tweet actions through Xquik.

Protect local AI coding-agent harnesses before tools run, review approvals and evidence, and scan agent plugins, skills, MCP servers, and marketplace packages with HOL Guard.

Turns folders, codebases, websites, PDFs, documents, and research topics into source-grounded HTML documents, deterministic narrated videos, or complete learning courses; also publishes and revises collaborative HTML pages through the HTML Docs CLI, REST API, and MCP server.

Turn Claude Code, Cursor, or Grok Bot into a 7-role Hyperliquid trading desk. Sixteen SKILL.md skills plus seven agent prompts cover market data, risk limits, ticketed execution, and post-trade review. You approve every trade by ticket id. Trade-only API wallet; no withdraw.

Use ImagineVid's OAuth-protected MCP and CLI tools to discover and safely run current image, video, and music generation capabilities.