Search
Find your next AI superpower — a-gnt helps you discover tools you didn't know existed.
53 results for “bench”
From the Community
Your Tools, Your Skills, Anywhere You Use AI — Across Providers
Your .claude folder is locked to one project on one machine. Your a-gnt library isn't. Here's how to call your skills, souls, and benches from any AI client that speaks MCP — and from OpenAI too.
blend-a-gnt Beta: a Work Task, a Bench, and a Prompt Walk Into a Bar
Sit down with a ticket, open blend-a-gnt, pick a bench, and see why mixing an MCP, a soul, and a prompt beats opening Claude cold.
The Spoonie Product Designer's Bench: Tools for Working With Chronic Illness in Tech
Chronic illness shapes how people work. AI tools can flatten the disability tax inside the work itself — not just in the products.
5 Soul Collections Every AI Fan Needs
Themed Soul collections you can save to your bench — from productivity powerhouses to creative chaos.
Benches
The Sole Proprietor's Bench
Twelve tools for a one-person business in 2026, picked the way a good shop neighbor picks them: by what actually works on a Monday morning. The six MCP servers connect Claude to the SaaS you already pay for — Shopify for the storefront, Zapier for the glue between everything else, HubSpot for the tiny pipeline, Asana for the work that has to ship this week, Canva for the flyer that's due tomorrow, Webflow for the marketing site nobody else will touch. The expense sorter handles the Sunday-night receipt pile before you hand the shoebox off. The desk attendant is for the moment you sit down and the inbox looks like a wall — you needed a calm voice, not another productivity system. Most one-person businesses don't fail because the owner couldn't think; they fail because the owner ran out of patience to do one more thing alone. This bench is what 'one more thing' looks like when you have help that actually understands the work.
7 tools · by a-gnt Community
Discoverable by Machines: The SEO/AEO Stack
Every piece of infrastructure you need to make a website findable by Google, ChatGPT, Claude, Perplexity, and the long tail of autonomous agents that are now deciding what humans read. This bench pairs the companion article ("The Whole Stack of Being Found") with real, working catalog tools: Search Console MCP servers for measuring what's already working, the AEO CLI for automating Answer Engine audits, the llms.txt hub for discovery, structured-data and crawling tools for ingestion, and the Claude SEO MCP for hands-on analysis inside your agent. Use it as a toolkit: install the Search Console MCP first so Claude or Cursor can pull your own data as tools, then run the AEO CLI over your site to generate the audit, then steal ideas from llms-txt-hub for your own manifest. The article at /blog/the-whole-stack-of-being-found walks through all fourteen pieces of the stack. The bench gives you the tools to actually ship them. SEO is not dead. It just grew a second half called AEO. Both matter. This bench is the shortest path I know to doing both.
13 tools · by a-gnt Community
For Your ADHD Brain, Specifically
Ten tools in this bench, chosen for the specific neurology you actually have. Not productivity hacks. Not focus apps. These are built around the patterns that real ADHD brains run on — the dropped thread when you opened the tab, the racing mind at 3 am, the hour that felt like five minutes, the task you've been avoiding for three days, the hyperfocus you couldn't break, the meltdown you couldn't see coming. Every one of them refuses to lecture you about time management. Every one of them knows that "just try harder" is not advice. Use them at the worst moments of the day, or the best. They'll be the same either way, which is the point.
10 tools · by a-gnt Community
The Late Bloomer's Starter Pack
It's never too late to start something. But starting in your 40s, 50s, or beyond comes with friction the early-bloomers don't have to think about — the embarrassment of being a beginner as an adult, the impatience to be good fast, the well-meaning friends who don't quite get it. This bench is for the person picking up an instrument, a language, a pen, a sketchbook, or a side hustle later in life. A blunt-warm mentor, a brutally kind hobby coach, a 30-day skill sprint planner, an honest side-hustle advisor, and a résumé rewriter for the career pivot you're flirting with. Five tools, one quiet permission slip.
5 tools · by a-gnt Community
The Caregiver's Toolkit
If you're caring for an aging parent, a chronically ill partner, or a kid with complex needs, the paperwork alone will eat your life. This bench gathers the tools that take the load off your shoulders without pretending to be a lawyer or doctor: a soul who decodes bureaucratic letters, a prompt that builds your 'what to ask the doctor' list, a skill for debriefing after appointments, an elder paperwork decoder, and a memory MCP server so your AI can finally hold context across the months-long conversation that caregiving really is. Pair them with a cup of coffee and the stack of unopened envelopes you've been avoiding.
5 tools · by a-gnt Community
The Content Designer's Microcopy Toolkit
The smallest words in your product carry the most weight. The button label that fails. The error message that shames. The empty state that gives up. This bench is for the content designer, UX writer, or technical writer who wants to ship microcopy that earns its place. A salty editor soul who refuses to let you ship 'an unexpected error occurred', three prompts for rewriting product strings, plain language passes, and error message remediation, and a cognitive load skill that flags every word doing too much. Run them on a single product flow and watch the cognitive load drop without a single design change.
6 tools · by a-gnt Community
Tools
Mcp Gateway
Universal MCP Gateway - Single-port multiplexing with Meta-MCP for ~95% context token savings
gabrielmaialva33/winx-code-agent
gabrielmaialva33/winx-code-agent
Caveman
🪨 why use many token when few token do trick — Claude Code skill that cuts 65% of tokens by talking
AgentBench
A Comprehensive Benchmark to Evaluate LLMs as Agents (ICLR'24)
Khoj
Your AI second brain. Self-hostable. Get answers from the web or your docs. Build custom agents, sch
XGenerationLab/xiyan_mcp_server
XGenerationLab/xiyan_mcp_server
Cc Switch
A cross-platform desktop All-in-One assistant tool for Claude Code, Codex, OpenCode, openclaw & Gemi
Mempalace
The highest-scoring AI memory system ever benchmarked. And it's free.
Code Mode
🔌 Plug-and-play library to enable agents to call MCP and UTCP tools via code execution.
Dbhub
Zero-dependency, token-efficient database MCP server for Postgres, MySQL, SQL Server, MariaDB, SQLit
Cua
Open-source infrastructure for Computer-Use Agents. Sandboxes, SDKs, and benchmarks to train and eva
AReaL
Lightning-Fast RL for LLM Reasoning and Agents. Made Simple & Flexible.
LLM Agents Papers
A repo lists papers related to LLM based agent
LLMCompiler
[ICML 2024] LLMCompiler: An LLM Compiler for Parallel Function Calling
Code Act
Official Repo for ICML 2024 paper "Executable Code Actions Elicit Better LLM Agents" by Xingyao Wang
Awesome GUI Agent
💻 A curated list of papers and resources for multi-modal Graphical User Interface (GUI) agents.
Agent Framework
A framework for building, orchestrating and deploying AI agents and multi-agent workflows with suppo
Agentops
Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most
Browser Use
🌐 Make websites accessible for AI agents. Automate tasks online with ease.
Claw Compactor
14-stage Fusion Pipeline for LLM token compression — reversible compression, AST-aware code analysis
Betterleaks
A Better Secrets Scanner built for configurability and speed
Cursor Rules Java
A curated collection of Skills and Agents for Java Enterprise development, designed to streamline mo
Phoenix
AI Observability & Evaluation
Workbench
From Idea to Execution - Manage your trading operation across a distributed cluster
PaddleOCR
Turn any PDF or image document into structured data for your AI. A powerful, lightweight OCR toolkit
Qdrant
Qdrant - High-performance, massive-scale Vector Database and Vector Search Engine for the next gener
Memvid
Memory layer for AI Agents. Replace complex RAG pipelines with a serverless, single-file memory laye
LEANN
[MLsys2026]: RAG on Everything with LEANN. Enjoy 97% storage savings while running a fast, accurate,
Infinity
The AI-native database built for LLM applications, providing incredibly fast hybrid search of dense
Webclaw
Fast, local-first web content extraction for LLMs. Scrape, crawl, extract structured data — all from
Browser
Lightpanda: the headless browser designed for AI and automation
BrowserGym
🌎💪 BrowserGym, a Gym environment for web task automation
Browserable
Open source and self-hostable browser automation library for AI agents
jsdelivr/globalping-mcp-server
jsdelivr/globalping-mcp-server
Daichi-Kudo/llm-advisor-mcp
Daichi-Kudo/llm-advisor-mcp
Whatsonyourmind/oraclaw
Whatsonyourmind/oraclaw
fabio-rovai/open-ontologies
fabio-rovai/open-ontologies
hifriendbot/cogmemai-mcp
hifriendbot/cogmemai-mcp
wazionapps/nexo
wazionapps/nexo
Langfuse
🪢 Open source LLM engineering platform: LLM Observability, metrics, evals, prompt management, playg
Asana Workbench
The Friday status update goes from an hour to a two-minute conversation.
Mini Swe Agent
The 100 line AI agent that solves GitHub issues or helps you in your command line. Radically simple,
Ruflo
🌊 The leading agent orchestration platform for Claude. Deploy intelligent multi-agent swarms, coord
realwigu/mcp-doctor
realwigu/mcp-doctor
ajitpratap0/GoSQLX
ajitpratap0/GoSQLX
gpartin/WaveGuardClient
gpartin/WaveGuardClient
g1itchbot8888-del/agent-memory
g1itchbot8888-del/agent-memory
Himalayas-App/himalayas-mcp
Himalayas-App/himalayas-mcp
JARVIS
JARVIS
Escape Room: The Clockmaker's Workshop
Solve mechanical puzzles to escape a Victorian workshop before time runs out
SWE Agent
SWE Agent
Courtroom Drama
You're a lawyer — argue cases with evidence, objections, and jury verdicts
Onyx
Open Source AI Platform - AI Chat with advanced features that works with every LLM