Best GitHub AI Agents & LLM Apps projects for SaaS builders

infiniflow/ragflow

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

iOfficeAI/AionUi

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

elizaOS/eliza

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

dataelement/bisheng

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

EverMind-AI/EverOS

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

NousResearch/hermes-agent

Offer a managed Hermes Cloud: hosted agent workspaces, persistent sandboxes, encrypted memory, multi-channel gateway, scheduled automations, team sharing, observability, backups, enterprise controls, and one-click deployment of skills/tools.

AI power users, developers, solo founders, research teams, devtool-heavy startups, and companies wanting persistent personal or team AI agents without building orchestration infrastructure.

trycua/cua

Hosted agent-ready computer sandboxes with APIs for screen, mouse, keyboard, shell, mobile gestures, dataset capture, benchmark runs, and fleet management across OS types.

AI agent startups, enterprise automation teams, QA automation teams, model labs, developer-tool companies, and research teams training or evaluating computer-use models.

makecindy/cindy

Sell the hosted coordination layer: sync, encrypted memory, team skills, scheduled automations, device handoff, plugin marketplace, managed MCP connectors, usage governance, audit logs, and cross-agent orchestration while keeping the client open source.

Developers, indie hackers, small engineering teams, AI-heavy operators, and technical founders who already pay for coding agents but want one persistent workspace across tools, devices, models, and automations.

feder-cr/AIHawk

Managed AI browser automation workspace with hosted browsers, reusable task templates, proxy/profile management, audit logs, retries, scheduling, human-in-the-loop approvals, and integrations with CRMs, Sheets, Slack, and internal tools.

Growth teams, recruiting agencies, sales ops, market research firms, SMB operators, and AI tool builders needing browser-based automation where APIs are missing or limited.

omnigent-ai/omnigent

Hosted Omnigent Cloud: managed agent sessions, browser/mobile access, cloud sandboxes, team collaboration, policy governance, usage controls, audit logs, SSO, role-based permissions, shared templates, and integrations with GitHub, Slack, Jira, and cloud providers.

Engineering teams, AI platform teams, CTOs, DevOps leaders, and developer productivity teams at startups and mid-market software companies adopting Claude Code, Codex, Cursor, OpenCode, or custom AI agents.

Windy3f3f3f3f/how-claude-code-works

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

kirodotdev/KiroCrew

Offer managed hosted workspaces, team collaboration, secure remote agents, audit logs, shared skills, scheduled automations, integrations, and enterprise controls on top of the open-source local runtime.

Engineering teams, startup founders, DevOps teams, platform engineering leads, and AI-forward software agencies.

CopilotKit/CopilotKit

Managed CopilotKit Cloud: hosted agent runtime, conversation/session state, analytics, debugging, user feedback loops, evals, auth/RBAC, enterprise governance, deployment templates, and premium integrations for React, Angular, Vue, mobile, and Slack.

Product engineering teams, AI-native SaaS startups, enterprise platform teams, developer tools teams, and companies adding embedded copilots or agent workflows to existing apps.

amitshekhariitbhu/ai-engineering-interview-questions

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

shy3130/tickflow-stock-panel

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

CherryHQ/cherry-studio

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

jo-inc/camofox-browser

Offer a hosted browser-agent runtime with managed Camoufox instances, residential proxy routing, session persistence, VNC login, trace/debug tools, API keys, concurrency scaling, uptime guarantees, and integrations for LangChain, OpenAI tools, MCP, Playwright, and Puppeteer-like workflows.

AI agent companies, web automation startups, data extraction teams, growth/research teams, QA automation vendors, and developers building browser-based agents that need reliable access to real websites.

spider-rs/spider

Managed crawling API and cloud platform that lets developers prototype locally with the open-source Rust crawler, then switch to hosted infrastructure for proxy routing, JS rendering, unblockers, distributed crawling, usage tracking, and reliability guarantees.

AI agent builders, LLM app developers, data engineering teams, growth/intelligence teams, and companies that need reliable web data extraction without maintaining crawler infrastructure.

the-open-engine/zeroshot

Hosted executor-verifier orchestration with team dashboards, policy controls, repo integrations, run history, CI evidence, approval workflows, cost tracking, and benchmarked agent/provider performance across tasks.

Engineering managers, platform engineering teams, AI tooling leads, CTOs at startups and mid-market SaaS companies using AI coding agents in real repositories.

QwenLM/qwen-code

Hosted control plane for Qwen Code teams: centralized policy management, model/provider routing, secrets isolation, usage and cost analytics, reusable team skills, MCP server management, audit trails, and managed remote agent workspaces.

Engineering teams, platform engineering leads, DevOps teams, AI tooling leads, and software agencies adopting coding agents across multiple developers.

langgenius/dify

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

synthetic-sciences/openscience

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

SylphxAI/pdf-reader-mcp

Hosted document-evidence API and team workspace for reliable PDF ingestion, table/OCR extraction, citation verification, trust reports, batch processing, monitoring, and integrations with agent/RAG stacks.

AI agent builders, legaltech teams, fintech/insurance analysts, enterprise RAG teams, compliance/document-review platforms, developer-tooling teams using MCP or SDK workflows.

can1357/oh-my-pi

A managed control plane for AI coding agents: centralized policy, model/provider routing, team workspaces, audit trails, secure secrets handling, repo-level permissions, shared agent sessions, usage analytics, and CI/code-review integrations built around the open-source CLI.

Engineering teams, AI-forward startups, platform engineering leads, developer productivity teams, and security-conscious companies adopting AI coding agents across multiple repositories.

vivekchand/clawmetry

Hosted ClawMetry Cloud for team-wide AI agent observability: centralized traces, token/cost analytics, runtime inventory, budget alerts, approval workflows, audit logs, policy controls, and integrations with Slack, PagerDuty, GitHub, and SSO.

Engineering teams, AI platform teams, DevOps leads, security-conscious startups, and enterprises adopting Claude Code, Codex, Cursor, Copilot, n8n, or other agent runtimes.

hugohe3/ppt-master

Hosted document-to-deck platform with editable PPTX export, brand templates, team workspaces, version history, enterprise templates, privacy controls, narration, and integrations with Google Drive, Notion, Slack, and PowerPoint.

Consultants, founders, sales/marketing teams, corporate strategy teams, educators, agencies, and knowledge workers who regularly turn documents into client-ready decks.

Canner/WrenAI

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

aipoch/open-science

Offer a managed research workspace with shared projects, versioned agent runs, governed model access, hosted scientific connectors, compliance-ready audit logs, team permissions, cloud execution, and premium domain-specific research skill packs.

Biotech startups, academic labs, computational biology teams, pharma R&D groups, CROs, and research operations teams that need AI-assisted analysis with reproducibility and data control.

Gitlawb/zero

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

gmickel/flow-next

Hosted agentic engineering control plane: persistent specs, workflow runs, cross-model review logs, receipts, team dashboards, policy gates, PR evidence, audit trails, and integrations with GitHub/GitLab/Slack/Linear/Jira.

Engineering managers, platform engineering teams, AI-forward software teams, CTOs at startups and mid-market SaaS companies using coding agents in production workflows.

Q00/ouroboros

Hosted control plane for agentic coding workflows: workflow registry, spec templates, run history, evaluation dashboards, policy enforcement, budget controls, team permissions, CI/GitHub integration, and compliance-grade audit logs across multiple coding-agent runtimes.

Engineering managers, platform teams, AI tooling leads, developer productivity teams, and startups using Claude Code/Codex/Gemini-style coding agents in production workflows.

asheshgoplani/agent-deck

Hosted AI-agent operations dashboard: team workspace for agent session monitoring, cost tracking, approvals, repo/worktree orchestration, shared transcripts, alerts, policies, and integrations with GitHub, Slack, Linear, Jira, and MCP servers.

AI-heavy engineering teams, indie hackers running many coding agents, devtools startups, platform/infra leads managing Claude Code/Codex/Gemini/OpenCode usage across repos.

moltis-org/moltis

Offer Moltis Cloud Control Plane: managed updates, encrypted remote access, backups, audit logs, policy management, team permissions, fleet monitoring, connector management, and compliance-ready deployment templates while keeping agent execution on customer hardware.

AI-native startups, developer teams, security-conscious SMBs, solo operators running personal agents, and enterprises experimenting with private AI assistants on owned hardware.

xalgord/xalgorix

Hosted autonomous pentest workspace with scheduled scans, verified exploit evidence, compliance-ready reports, team workflows, asset inventory, risk prioritization, and integrations with Jira, Slack, GitHub, and cloud environments.

Security teams at SMBs and mid-market SaaS companies, managed security service providers, bug bounty teams, and compliance-driven startups that cannot afford frequent manual pentests.

openwong2kim/wmux

Turn wmux into a subscription-grade AI agent operations workspace: team sync, shared agent sessions, cloud-backed durable workspaces, audit logs, policy controls, prompt/run history, managed templates, usage analytics, and secure approval workflows for multi-agent coding.

AI-heavy software engineers, indie hackers, engineering leads, devtool power users, and small teams adopting multiple coding agents in daily development.

webbrain-one/webbrain

Offer a hosted WebBrain Cloud subscription with reliable default models, saved workflows, task history, team templates, secure permissions, cross-browser sync, managed automation credits, and integrations with Google Sheets, Notion, Airtable, CRMs, and internal tools.

Knowledge workers, growth teams, sales/research teams, recruiters, analysts, solo founders, and small businesses that spend hours doing repetitive browser-based research or form-heavy workflows.

artokun/comfyui-mcp

Managed ComfyUI agent workspace: hosted GPU environments, natural-language workflow editing, model/workflow packs, team presets, job history, versioned workflows, collaboration, and support around reliable media generation.

AI creators, indie studios, prompt engineers, marketing/design teams, ComfyUI power users, small agencies producing images/video/audio with open models.

neomjs/neo

Hosted AI engineering cockpit that connects to GitHub repos and live apps, builds a persistent knowledge graph, coordinates multiple coding/review agents, detects runtime/UI issues, opens PRs, and tracks self-improving maintenance loops.

CTOs, VP Engineering, platform engineering teams, frontend infrastructure teams, and AI-forward software agencies maintaining complex web applications.

Hmbown/Codewhale

Hosted team workspace for Codewhale: centralized policy management, shared skills/agent roles, approval rules, MCP configuration, session history, cost analytics, compliance logs, and fleet orchestration across developer machines.

Engineering managers, platform teams, DevOps leads, and security-conscious startups using AI coding agents across multiple developers.

eugeniughelbur/obsidian-second-brain

Managed AI memory workspace that syncs with Obsidian/GitHub/YouTube/web/X, deduplicates and reconciles notes, maintains project memory, generates codebase docs, and gives teams searchable AI-ready context across tools.

AI-heavy founders, solo operators, researchers, developer teams, agencies, and small teams using Claude Code/Codex/Gemini plus Obsidian or markdown-based knowledge bases.

whiteguo233/OpenBiliClaw

hosted vertical AI agent, RAG workspace, or agent evaluation platform

AI teams, agencies, support teams, and ops-heavy SMBs

modu-ai/moai-adk

Turn the open-source harness into a hosted agent governance and observability platform: usage dashboards, cost budgets, policy enforcement, quality gate reporting, team templates, audit logs, and managed workflow orchestration.

Engineering managers, platform engineering teams, AI developer tooling teams, CTOs at startups using Claude Code or multi-agent coding workflows.

1jehuang/jcode

Turn the lightweight local harness into a hosted control plane for multi-agent coding workflows: session orchestration, shared memory, provider routing, team policies, usage/cost analytics, secure MCP/tool permissions, audit logs, and CI/PR integration.

Engineering teams, AI-native startups, platform/DevEx teams, CTOs managing multiple AI coding agents, and enterprises experimenting with Claude/OpenAI/Codex-style coding workflows.

addsumtech/slides_maker

Cloud workspace for source-traced deck generation with upload/import, branded templates, revision history, citation checking, critic review, team collaboration, export to PPTX/Keynote/PDF, and integrations with Google Drive, Notion, GitHub, arXiv, and Slack.

Researchers, consultants, startup founders, product marketers, analysts, educators, and technical teams who frequently turn papers, repos, docs, or data into client- or stakeholder-ready presentations.

santifer/career-ops

Hosted job-search copilot with job-board ingestion, fit scoring, resume tailoring, application CRM, alerts, analytics, and coach/team workflows, removing local CLI setup friction.

Job seekers in competitive markets, career switchers, laid-off tech workers, bootcamp graduates, and premium career coaches serving multiple clients.

qufei1993/skills-hub

Turn the desktop skills manager into a cloud-backed SkillsOps platform: private skill registry, team workspaces, cross-device sync, approved skill catalogs, version pinning, policy enforcement, usage analytics, and managed rollout to multiple AI coding tools.

AI-heavy engineering teams, developer productivity leads, platform engineering teams, agencies managing multiple AI coding tools, and power users using Claude Code/Codex/Cursor/OpenCode across projects.

BlockRunAI/Franklin

Offer a managed control plane for autonomous economic agents: wallet funding, prepaid balance, spend policies, model/tool routing, trading journal, approval workflows, team permissions, execution logs, and marketplace access to paid APIs/data sources.

Crypto-native traders, on-chain analysts, AI power users, trading communities, fintech builders, and teams that want agents to manage paid API/model/tool usage under budget controls.

LazyAGI/LazyLLM

Managed LazyLLM Cloud: hosted app builder, agent workflow templates, one-click deployment, model/provider switching, evaluation dashboards, bad-case analysis, fine-tuning pipelines, collaboration, versioning, and enterprise deployment controls.

AI engineering teams, enterprise innovation teams, ML platform teams, agencies building LLM apps for clients, and startups needing faster multi-agent/RAG deployment.

ZhuLinsen/daily_stock_analysis

Hosted AI stock research workspace with managed data connectors, scheduled reports, watchlist monitoring, explainable signal dashboards, team alerts, historical reports, and integrations with Telegram, Slack, Discord, email, and enterprise chat tools.

Retail investors, small investment clubs, financial content creators, independent advisors, and small funds that need automated market monitoring across US, China, Hong Kong, Japan, Korea, Taiwan stocks and ETFs.

simonlin1212/TradingAgents-astock

Hosted A-share AI research workbench that runs multi-agent reports, tracks policy/news/lockup/hot-money signals, stores watchlists, compares bull/bear theses, and exports structured investment research notes with compliance disclaimers.

Retail quant investors, China-focused independent analysts, small investment research teams, fintech educators, and developers building A-share research tools.

xpf0000/FlyEnv

A paid Pro/Team layer for synced project environments, team templates, AI-agent workspace profiles, shared service recipes, secure MCP permissions, backups, diagnostics, and onboarding automation across Windows/macOS/Linux.

Indie developers, web agencies, Laravel/PHP/Node/Python teams, full-stack product teams, and engineering managers who want simpler local setup across OSes without Docker overhead.

aiworkskills/wechat-article-skills

Turn the open-source local skill pack into a hosted subscription platform for brand configuration, team workflows, content calendars, templates, approvals, analytics, asset management, and one-click export/publish support.

Chinese creators, small media teams, founders, agencies, content marketers, and companies running WeChat Official Accounts.

esengine/DeepSeek-Reasonix

Build a team control plane for Reasonix: centralized agent sessions, policy management, run history, checkpoints, approvals, cost tracking, model/provider routing, workspace templates, and compliance logs.

Engineering teams, AI-native startups, platform engineering leads, developer productivity teams, and agencies running long-lived coding agents across multiple repos.

rcarmo/piclaw

Offer a managed PiClaw Cloud or enterprise control plane with hosted workspaces, backups, user management, audit logs, secure secrets, prebuilt add-ons, team sharing, usage analytics, and one-click local/VPC deployment.

Small engineering teams, AI-heavy agencies, internal developer-platform teams, security-conscious startups, and consultants running private LLM/coding-agent workspaces.

thClaws/thClaws

Offer thClaws Cloud or Team Console: managed workspace sync, shared agent memory, team permissions, audit logs, provider key management, hosted runners, connector marketplace, usage/cost tracking, policy controls, and one-click deployments for Slack/Discord/Telegram/LINE agents.

AI-forward engineering teams, developer tool teams, agencies, and small companies that want self-hostable AI agents across desktop, CLI, chat apps, and internal workflows without building their own harness.

agentforce314/clawcodex

Offer a hosted ClawCodex Cloud dashboard that manages team seats, provider credentials, model routing, prompt-cache optimization, audit logs, policy controls, shared agent profiles, usage analytics, and secure remote execution for coding agents.

Engineering managers, AI platform teams, DevTools startups, agencies, and cost-conscious software teams already using Claude Code/Codex-style coding agents.

openocta/openocta

Offer a cloud control plane for managing desktop agents, encrypted sync, team policies, skill marketplace subscriptions, audit logs, remote deployment, enterprise connectors, and managed model/provider configuration.

IT managers, DevOps teams, SMB operators, productivity-heavy professionals, and Chinese companies needing local-first AI automation with WeChat/DingTalk/Feishu integration.

tigicion/dao-code

Offer a managed Dao Code cloud/workspace subscription with team memory, repo indexing, secure execution sandboxes, usage analytics, cost controls, hosted DeepSeek routing, eval dashboards, shared skills, policy controls, and enterprise audit logs.

Cost-sensitive software teams, indie hackers, Chinese developers, AI-native startups, agencies, and engineering managers who want Claude-Code-like workflows without Anthropic access, quota friction, or high inference cost.

clawdotnet/openclaw.net

Managed AgentQi/OpenClaw control plane for monitoring, policy, approvals, evidence bundles, workflow governance, MCP app lifecycle, team access, deployment templates, and compliance reporting across self-hosted OpenClaw.NET instances.

Engineering teams, platform teams, .NET-heavy enterprises, regulated SMBs, and AI automation consultancies deploying internal agents or customer-facing agent workflows.

sonichi/sutando

Offer a managed Sutando Pro layer: packaged Mac app, secure setup, hosted sync, memory, skill marketplace, meeting/call automation, observability, approvals, backups, and team/admin controls while letting users bring their own LLM CLI/API credentials.

AI power users, solo founders, developers, consultants, and small teams who already use Claude Code/Codex and want a voice-driven personal automation layer across desktop, meetings, calls, and coding workflows.

zhu1090093659/deepseek-pp

Offer a managed DeepSeek productivity workspace around the extension: encrypted cloud sync, cross-device memory, team prompt/skill libraries, automation scheduler, managed MCP connectors, audit logs, exports, and admin controls.

DeepSeek power users, AI-heavy freelancers, researchers, developers, small teams, and China/Asia-focused companies that rely on DeepSeek Web but need workflow persistence, export, memory, browser automation, and MCP tooling.

EvoScientist/EvoScientist

A hosted AI research co-pilot platform with persistent lab memory, autonomous literature monitoring, hypothesis pipelines, citation-backed reports, project workspaces, benchmarked research agents, integrations with arXiv/PubMed/Zotero/Notion/GitHub, and team collaboration controls.

AI/biotech research labs, university PI teams, corporate R&D teams, scientific founders, technical due-diligence teams, and grant-writing/research strategy consultants.

NotASithLord/peerd

Offer a hosted management layer for peerd: team policy controls, encrypted credential orchestration, audit logs, approved automation templates, browser fleet deployment, shared agent workspaces, usage analytics, and enterprise security/compliance features.

AI-forward product, operations, QA, research, and growth teams that want agents to work directly inside browser workflows without building custom automation infrastructure.

rlaope/oh-my-hermes

Turn OMH into a hosted team control plane for AI-agent workflows: shared workflow libraries, governed runbooks, audit trails, reusable skill packs, team memory, policy controls, and analytics across agent runs.

AI power users, solo developers, small engineering teams, AI-agent consultants, and teams already using Hermes Agent, Claude Code, Codex, or similar coding-agent workflows.

ArtificialAnalysis/Stirrup

Turn Stirrup into a managed agent operations platform: hosted agent runtime, secure tool execution, MCP connector management, eval suites, trace replay, cost monitoring, versioned skills, prompt/context debugging, and team permissions.

AI product teams, developer-tooling teams, startups building internal agents, consultancies delivering custom LLM agents, and enterprises that need controllable agent infrastructure.

TNT-Likely/PanWatch

Turn the self-hosted app into a hosted AI investment cockpit with managed data feeds, portfolio sync, scheduled AI reports, real-time alerts, historical signal tracking, and team/shared-watchlist features.

Active retail investors, high-net-worth individual traders, small investment clubs, independent advisors, and Asia-focused stock traders monitoring A-shares, Hong Kong stocks, and US equities.

he-yufeng/CoreCoder

Hosted learning and customization platform for building internal coding agents: interactive source walkthroughs, sandbox runs, templates, evaluation harnesses, policy controls, and team workspaces.

AI engineering leads, developer-tooling teams, AI bootcamps, technical educators, and startups building custom coding agents

xiaojilele-glitch/WhyBuddy

A collaborative AI product discovery platform: idea-to-PRD, market assumptions, user flows, architecture, prompt packs, prototype previews, review workflows, and export packages for teams building new SaaS products.

Startup founders, product managers, indie hackers, agencies, innovation teams, and AI-native software studios that need to validate and specify product ideas faster.

yashab-cyber/opendroid

Offer a subscription cloud layer around the open-source Android agent: managed workflow templates, remote configuration, audit logs, device fleet management, premium LLM routing, secure memory sync, team policies, and marketplace automations.

Mobile-first teams, field service operators, Android power users, customer support teams using WhatsApp/SMS, accessibility automation developers, and small businesses that rely on repetitive phone workflows.

gakonst/nanocodex

Offer a hosted agent runtime and control plane around Nanocodex: managed sandboxes, session persistence, evals, logs, team workspaces, policy controls, billing, observability, and deployment integrations.

AI-native developer-tool companies, internal platform teams, dev agencies building coding agents, and startups embedding autonomous coding workflows into their products.

code-yeongyu/senpi

Turn senpi into a hosted control plane for coding-agent runtimes: policy management, team permissions, audit logs, model presets, extension marketplace, usage analytics, and CI/IDE integration.

AI-native engineering teams, devtool startups, platform teams, and companies adopting coding agents in regulated or production codebases.

grinev/opencode-telegram-bot

Turn the bot into a managed remote agent operations layer: hosted bot setup, multi-device session dashboard, secure relay, team permissions, audit logs, notification rules, scheduled agent jobs, worktree/session management, and integrations with GitHub/Slack/Linear.

Solo developers, indie hackers, small engineering teams, AI-heavy builders, and technical founders who run coding agents locally but want mobile monitoring, approvals, notifications, and scheduled task control.

EvoScientist/EvoSkills

Turn the open-source skill packs into a managed research-agent workspace with curated premium skills, team memory, compliance controls, reproducible experiment pipelines, paper/project templates, evaluation dashboards, and managed MCP/tool integrations.

AI research labs, university research groups, biotech/chemistry/materials startups, and R&D teams using coding agents for literature review, experiments, paper writing, and reproducible research.