# AI Agents Info

> Find AI tools and agents in one place. Independent listings checked against a live site, plus original reporting on pricing, harnesses, and computer use.

The sections below are generated from the same files that render the site, so new posts, research notes, glossary terms, and directory listings appear here on the next deploy. Prices, model names, and revenue figures are dated on the page they come from. A private tool-candidate backlog is not published.

The longer plain-text copy of posts, research, glossary entries, and directory listings is at https://www.aiagentsinfo.com/llms-full.txt.

## Key sections

- [Blog](https://www.aiagentsinfo.com/blog): Explainers, news, and guides.
- [Research](https://www.aiagentsinfo.com/research): Paper breakdowns and original notes.
- [Papers](https://www.aiagentsinfo.com/papers): Paper breakdowns, each with the source link and the paper's authors.
- [Guides](https://www.aiagentsinfo.com/blog?category=guide): Blog posts filed as guides.
- [Glossary](https://www.aiagentsinfo.com/glossary): Short definitions used across the site.
- [Directory](https://www.aiagentsinfo.com/directory): AI tools, apps, SaaS, and agents, each with a source site.
- [AI agents directory MCP server](https://www.aiagentsinfo.com/mcp): Public, read-only Model Context Protocol server at /api/mcp for searching listings, comparisons, and research.
- [Live data trackers](https://www.aiagentsinfo.com/trackers): Download, AGENTS.md, and GitHub star counts that refresh daily from public APIs.
- [AI agent framework downloads tracker](https://www.aiagentsinfo.com/trackers/downloads): PyPI and npm file-request counts for major agent frameworks. A download is not a user.
- [AGENTS.md adoption tracker](https://www.aiagentsinfo.com/trackers/agents-md): GitHub code search for AGENTS.md files, next to the October 2026 census of 6,006 of 22,854 popular repos.
- [AI agent framework GitHub stars tracker](https://www.aiagentsinfo.com/trackers/stars): Live stargazers_count for the repositories behind those frameworks.
- [Breakthroughs timeline](https://www.aiagentsinfo.com/trackers/breakthroughs): Public agent moments, added only with a date and a source.
- [Best AI agents](https://www.aiagentsinfo.com/best): Sourced comparisons of harnesses, computer use, evaluation tools, and consumer jobs such as shopping, search, travel, email, and writing. Last verified November 2026.
- [Best AI agent harness](https://www.aiagentsinfo.com/best/ai-agent-harnesses): Sourced comparison of coding, open-source, and local agent harnesses. Last verified November 2026. No invented scores.
- [AI agent framework comparison table](https://www.aiagentsinfo.com/best/ai-agent-frameworks): Nineteen open-source agent frameworks and SDKs. Licence and latest GitHub release read November 9, 2026. Not a quality ranking.
- [Best AI coding agents](https://www.aiagentsinfo.com/best/ai-coding-agents): Four coding agents. List prices read November 9, 2026 from the vendor pages. Free and paid. Not a quality ranking.
- [Best computer use agents](https://www.aiagentsinfo.com/best/computer-use-agents): Sourced comparison of desktop, API, open-source, and Windows computer use agents. Last verified November 2026. No invented scores.
- [Best AI agent evaluation tools](https://www.aiagentsinfo.com/best/ai-agent-evaluation-tools): Sourced comparison of agent evaluation and trajectory-testing tools, including open source. Last verified November 2026.
- [Best AI agent for shopping](https://www.aiagentsinfo.com/best/ai-agents-for-shopping): Sourced comparison of shopping agents on checkout, seller, and list price. Last verified November 2026. No invented scores.
- [Best AI agent for web search](https://www.aiagentsinfo.com/best/ai-agents-for-web-search): Sourced comparison of web-search agents on plans, citations, and list price. Last verified November 2026.
- [Best AI agent for travel booking](https://www.aiagentsinfo.com/best/ai-agents-for-travel-booking): Sourced comparison of travel-booking agents on who takes payment and list price. Last verified November 2026.
- [Best AI agent for email](https://www.aiagentsinfo.com/best/ai-agents-for-email): Sourced comparison of email agents on inbox, plans, and list price. Last verified November 2026.
- [Best AI agent for personal assistant tasks](https://www.aiagentsinfo.com/best/ai-agents-for-personal-assistant-tasks): Sourced comparison of personal-assistant agents on scheduled work and list price. Last verified November 2026.
- [Best AI agent for research](https://www.aiagentsinfo.com/best/ai-agents-for-research): Sourced comparison of research agents on reports, notebooks, and list price. Last verified November 2026.
- [Best AI agent for scheduling](https://www.aiagentsinfo.com/best/ai-agents-for-scheduling): Sourced comparison of scheduling agents on who writes the event and list price. Last verified November 2026.
- [Best AI agent for writing](https://www.aiagentsinfo.com/best/ai-agents-for-writing): Sourced comparison of writing agents on where the draft lives and list price. Last verified November 2026.
- [Best AI agent for students](https://www.aiagentsinfo.com/best/ai-agents-for-students): Sourced comparison of student-facing agents on school rules, tutors, and list price. Last verified November 2026.
- [Best AI agent for small business](https://www.aiagentsinfo.com/best/ai-agents-for-small-business): Sourced comparison of small-business agents on seats, inboxes, and list price. Last verified November 2026.
- [Harness comparison](https://www.aiagentsinfo.com/trackers/harnesses): Pointer to the sourced harness comparison. Cells are published only with a source.
- [Benchmark leaderboard](https://www.aiagentsinfo.com/trackers/benchmarks): Public agent benchmarks, published only with a source, a date, and the harness.
- [Reports](https://www.aiagentsinfo.com/reports): Dated pages with a source and a citation.
- [Browser and computer-use agents next 6 months: 10 products, one date](https://www.aiagentsinfo.com/reports/browser-and-computer-use-agents-next-6-months): Ten live browser and computer-use agents from November 9, 2026 through May 9, 2027. The only named day is November 18, 2026. Checked November 9, 2026.
- [AI coding agents in the next 30 days (8 Nov to 8 Dec 2026)](https://www.aiagentsinfo.com/reports/ai-coding-agents-next-30-days): Prices and usage limits for 11 coding agents, plus three dated marks. Checked November 8, 2026.
- [AI agents in the next 12 months (Nov 2026 to Nov 2027)](https://www.aiagentsinfo.com/reports/ai-agents-next-12-months): Flagship outlook. Announced, likely, and watch-for claims for coding, computer use, harnesses, and regulation through November 2027. Read November 7, 2026.
- [AI agents in the next 6 months (Nov 2026 to May 2027)](https://www.aiagentsinfo.com/reports/ai-agents-next-6-months): Announced, likely, and watch-for claims for coding, computer use, harnesses, and regulation. Read November 5, 2026.
- [AI shopping agents next 6 months: 6 companies](https://www.aiagentsinfo.com/reports/ai-shopping-agents-next-6-months): What Google, OpenAI, Amazon, Microsoft, Shopify, and Adobe have dated from November 2026 through May 2027. Checked November 5, 2026.
- [Agent harness benchmark, November 2026](https://www.aiagentsinfo.com/reports/agent-harness-benchmark-november-2026): Agent harness benchmark for November 2026: public SWE-bench and Terminal-Bench scores for coding agents, each with the date on the official source.
- [AI agent framework release November 2026: sortable table of 19 repos](https://www.aiagentsinfo.com/reports/ai-agent-framework-github-stats-november-2026): AI agent framework release November 2026: a sortable table of the latest GitHub releases for 19 open-source frameworks and SDKs, pulled 5 November 2026. 14 of 19 were within 30 days. Stars, forks, licences, and release dates are on this page.
- [Verified revenue](https://www.aiagentsinfo.com/verified-revenue): Payment-processor-verified revenue. Checked 2026-11-05.
- [Directory stats](https://www.aiagentsinfo.com/directory/stats): Counts generated from the directory files at build time.
- [AI agent cost calculator: free 30-day bill](https://www.aiagentsinfo.com/agent-cost-calculator): Free AI agent and chatbot cost calculator. 21 official models. Prices read November 9, 2026.
- [Prompt and tool checker](https://www.aiagentsinfo.com/agent-prompt-checker): Server-side check of pasted prompts and OpenAI, Anthropic, or MCP tool JSON, with a token count and the same official prices.
- [Submit a tool](https://www.aiagentsinfo.com/submit): Public form. A valid submission is published to the directory.

## Latest blog posts

- [Agent tool JSON cost: $2.4672 to $17.136 for one list (Nov 9, 2026)](https://www.aiagentsinfo.com/blog/agent-tool-json-cost-across-models): 2026-11-09. On November 9, 2026 the prompt checker priced one three-tool support list at 200 output tokens and 40 turns a day. gpt-5, a tiktoken count, was $3.042 over 30 days. The same paste was $2.4672 on grok-4.7 and $17.136 on gpt-6-astra, both estimates.
- [Claude Code Don't Ask Again: 1,122 Public Files Held a Secret (Nov 2026)](https://www.aiagentsinfo.com/blog/claude-code-dont-ask-again): 2026-11-09. 1,122 public copies of .claude/settings.local.json held at least one secret, 1,379 in all, in the census The Hacker News and iTnews attribute to Apiiro. Claude Code writes that file when you pick Yes, and don't ask again.
- [How to Build an Eval Harness for AI Agents: Isolate First](https://www.aiagentsinfo.com/blog/how-to-build-an-eval-harness-for-ai-agents): 2026-11-09. How to build an eval harness for AI agents: write tasks with checkable outcomes, isolate every trial, grade the environment, and keep a held-out set. Harbor, Inspect, and a local pytest suite are three ways to run that loop.
- [How to Get Your AI Tool Recommended by ChatGPT and Google (2026 Founder Guide)](https://www.aiagentsinfo.com/blog/how-to-get-your-ai-tool-recommended-by-chatgpt-and-google): 2026-11-09. How to get your AI tool recommended by ChatGPT and Google: allow the search crawlers, publish unique pages they can cite, and treat a directory listing as a public URL — not a paid rank. OpenAI and Google both say eligibility is not a guarantee.
- [Is AutoGen Deprecated? 61,753 Stars on an Unarchived Repo](https://www.aiagentsinfo.com/blog/is-autogen-deprecated): 2026-11-09. 61,753 GitHub stars on November 9, 2026, and the API field archived was false. AutoGen’s README, changed October 16, says Microsoft is not providing bug fixes, security fixes, new features, or support.
- [Nvidia board investigation: Jensen Huang stays CEO while directors review the reports](https://www.aiagentsinfo.com/blog/nvidia-board-investigation-jensen-huang): 2026-11-09. On November 8, 2026, Nvidia’s board opened an independent investigation into reporting that Jensen Huang was the target of a foreign intelligence effort. The company says he remains CEO.
- [What Does Ultracode Do in Claude Code? It Plans a Workflow at Any Effort (2026)](https://www.aiagentsinfo.com/blog/what-does-ultracode-do-in-claude-code): 2026-11-09. Ultracode plans a dynamic workflow for each substantive Claude Code task, at the effort level the session already uses. Claude Code v2.1.284, dated September 28, 2026, is the release that stopped forcing that session to xhigh.
- [Multi-Agent Orchestration: Removing the Interaction Term Costs 7.2 to 17.3 Points on Code (2026)](https://www.aiagentsinfo.com/blog/what-is-multi-agent-orchestration): 2026-11-09. What is multi-agent orchestration? An orchestrator chooses the next worker and what that worker may read. In Liu and Xiong, arXiv November 5, 2026, dropping the interaction term costs 7.2 points on their long-input HumanEval and 17.3 points on MBPP.
- [Which AI Agent Framework Do People Install? 155,918,257 LangChain Downloads in 30 Days](https://www.aiagentsinfo.com/blog/which-ai-agent-framework-do-people-install): 2026-11-09. 155,918,257 PyPI downloads for langchain from October 10 through November 8, 2026, without mirrors. LangGraph had 55,093,579. CrewAI had 2,629,769 and 15,995 more GitHub stars than LangGraph.
- [How Do AI Agents Work? The Control Loop From Goal to Action (2026)](https://www.aiagentsinfo.com/blog/how-do-ai-agents-work): 2026-11-08. How do AI agents work? A language model proposes a step. A harness runs an allowed tool, returns the result, and the loop continues until the goal is met, a limit hits, or a person must approve.
- [How Many Official MCP Servers Are There? 59,506 Latest Records on November 8, 2026](https://www.aiagentsinfo.com/blog/how-many-official-mcp-servers-are-there): 2026-11-08. 59,506 latest records in the official MCP registry, counted from 13:43 to 13:45 UTC on November 8, 2026. 58,565 were marked active. One namespace held 2,366 of them.
- [What Is BioBasicsBench? How Well AI Agents Handle Everyday Wet-Lab Science](https://www.aiagentsinfo.com/blog/what-is-biobasicsbench): 2026-11-08. What is BioBasicsBench? It is OpenAI’s November 6, 2026 benchmark, built with Ligo, that scores AI agents on 100 raw wet-lab tasks. At the xhigh setting, GPT-6 Astra scored 66.7% of rubric credit.
- [What Is T3 Code? 51,402 GitHub Stars on an Open-Source Control Plane (2026)](https://www.aiagentsinfo.com/blog/what-is-t3-code): 2026-11-08. 51,402 GitHub stars at 03:46 UTC on November 8, 2026. T3 Code is a free MIT control surface for coding agents you already subscribe to, and release v0.1.7 shipped on November 4.
- [What is an AI employee? A digital worker, not a legal employee](https://www.aiagentsinfo.com/blog/what-is-an-ai-employee): 2026-11-07. What is an AI employee? Workplace software, also called a digital worker, that takes a goal, uses tools, and returns work for a person to review. Always-on AI agents, on the pages checked November 7, 2026, keep a cloud task running after the laptop closes or start on a schedule. The software is not a legal employee.
- [What Is Harness Engineering at Scale? Thoughtworks Radar Volume 35 Explained](https://www.aiagentsinfo.com/blog/what-is-harness-engineering-at-scale-thoughtworks-radar): 2026-11-07. What is harness engineering? It is the work of designing the software around an AI agent. Thoughtworks Technology Radar Volume 35, released November 4, 2026, makes harness engineering at scale a theme and treats the agent harness as a product.
- [Why Tool JSON Costs More Than Your System Prompt (and How to Check It)](https://www.aiagentsinfo.com/blog/why-tool-json-costs-more-than-your-system-prompt): 2026-11-07. On November 7, 2026 the prompt checker counted a 23-token weather prompt and a 94-token get_weather tool in OpenAI's Responses shape. For gpt-5, with 200 output tokens entered and 20 turns a day, the 30-day figure was $1.28775. The tool text was $0.0705 of that. The page rounds both this result and a 101-token Chat Completions paste to $1.29, so the token split is the number to read.
- [Claude Code vs Codex Pricing: Which Costs Less in 2026?](https://www.aiagentsinfo.com/blog/claude-code-vs-codex-pricing): 2026-11-06. Codex costs less to start. On November 6, 2026 it is included on ChatGPT Free and on Go at $8 a month, and Claude Code starts on Pro at $20 a month, or $17 a month on an annual bill. Anthropic Claude Code limits on Pro are a five-hour window and a weekly cap, shared with the rest of Claude. The $100 and $200 individual prices match, and Codex Pro also lists a $500 tier.
- [How to Evaluate an AI Agent Harness (Nov 2026)](https://www.aiagentsinfo.com/blog/how-do-you-evaluate-an-ai-agent-harness): 2026-11-06. How to evaluate an AI agent harness: test one action you can see, then check the whole job still finishes. A benchmark score does not name which step moved.
- [Can an AI agent use your computer? Yes (checked Nov 7, 2026)](https://www.aiagentsinfo.com/blog/can-an-ai-agent-use-your-computer): 2026-11-05. Yes, when the product has computer use. On November 7, 2026, Claude and ChatGPT desktop apps and the Claude, OpenAI, and Gemini APIs can click, type, and scroll.
- [Tencent releases Hy4 with image understanding and computer use](https://www.aiagentsinfo.com/blog/tencent-releases-hy4-with-computer-use): 2026-11-05. On November 4, 2026, Tencent released Hy4, a model that can read images and operate a computer. The API is live. On November 9 the announcement still had not published the weights.

## Research

- [Garry Tan Harness Evals: $0.17 to $0.44 per Solved Task (Nov 2026)](https://www.aiagentsinfo.com/research/garry-tan-harness-evals-november-2026): 2026-11-09. Seven harness cells on November 5, 2026 scored 71.3% to 78.8% on 20 tasks, and all six preregistered pass-rate tests came back can't tell apart. OpenClaw on Opus 5.5 cost $0.4435 per solved task; Claude Code cost $0.2170.
- [How Many GitHub Stars Does OpenClaw Have? 394,222 on November 9, 2026](https://www.aiagentsinfo.com/research/how-many-github-stars-does-openclaw-have): 2026-11-09. 394,222 stars on openclaw/openclaw at 07:41 UTC on November 9, 2026. Hermes Agent had 261,856. Claude Code had 155,043, and its LICENSE.md reserves all rights.
- [How Many Repos Use AGENTS.md? 6,006 of 22,854 on October 16, 2026](https://www.aiagentsinfo.com/research/how-many-repos-use-agents-md): 2026-11-09. 6,006 of 22,854 popular, active GitHub repositories had a root AGENTS.md or CLAUDE.md on October 16, 2026. That is 26.3%. The AGENTS.md homepage, read November 9, still says over 60,000 projects.
- [Parallel agent teams beat the best of 100 solo agents on 10 of 12 tasks (2026)](https://www.aiagentsinfo.com/research/parallel-agent-teams-test-time-compute): 2026-11-09. November 6, 2026 arXiv paper, in the November 9 listing: with Qwen3.8-27B, ten teams of ten beat the expected best of 100 solo agents on 10 of 12 tasks.
- [Paper breakdown: MCPHallu — why MCP agents fail on unreachable goals](https://www.aiagentsinfo.com/research/mcphallu-mcp-agent-hallucinations): 2026-11-08. Average score 0.642 across 14 agents and 5,012 trajectories. Unreachable Goal scores 0.458, and 23.2% of those runs pass. Claude Opus 4.6 leads overall at 0.728. No model in the paper scores above that.
- [Terminal-Bench 2.1: 28 of 89 repaired (2026)](https://www.aiagentsinfo.com/research/terminal-bench-2-1-benchmark-repairs): 2026-11-08. Terminal-Bench 2.1 repairs 28 of 89 tasks from version 2.0. Individual scores shift by up to 12 points, ranks move by at most 3. Spearman 0.958.
- [Evolver agent exploration: 320 runs, 83.13% to 87.50%](https://www.aiagentsinfo.com/research/evolver-agent-exploration-evolving-capabilities): 2026-11-07. Evolver agent exploration, arXiv November 5, 2026: 320 paired diagnostic runs inside Claude Code and Codex. Success rose from 83.13% to 87.50%. Recorded task-execution tokens fell 24.24%. This page has the tables.
- [HarnessFlow MCTS harness optimization: 69.48% average](https://www.aiagentsinfo.com/research/harnessflow-mcts-harness-optimization): 2026-11-06. HarnessFlow MCTS harness optimization, arXiv November 4, 2026: with 15 refinement attempts per benchmark, the average is 69.48%, 10.89 points above Vanilla and 3.95 points above Meta-Harness. The comparison table is on this page.
- [SpecCompact agent context compaction: 8.68× median](https://www.aiagentsinfo.com/research/speccompact-agent-context-compaction): 2026-11-05. SpecCompact agent context compaction, posted November 4, 2026: 2.04–5.63× on AppWorld and τ³, and an 8.68× median in a Hermes user study. The speed tables are on this page.
- [Paper breakdown: You don't need long context — short working memory still finishes the job](https://www.aiagentsinfo.com/research/you-dont-need-long-context): 2026-11-04. A November 3, 2026 technical report: compaction kept Claude Opus 4.5 and GPT 5.6 Sol near full-history accuracy while the live context stayed short.
- [Paper breakdown: Weight-space programming — a language model that reads and edits neural weights](https://www.aiagentsinfo.com/research/weight-space-programming): 2026-11-03. MIT CSAIL note, November 2, 2026: GPT-6 Astra edited neural weights, with no training loop, and raised a small model's MATH-500 score from 7.9% to 48.1%.
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://www.aiagentsinfo.com/research/react-reasoning-and-acting): 2026-10-18. ReAct (Yao et al., 2022) interleaves a reasoning trace with an action, then reads the observation. The paper reports this on HotpotQA, FEVER, ALFWorld, and WebShop.

## Guides

- [Agent tool JSON cost: $2.4672 to $17.136 for one list (Nov 9, 2026)](https://www.aiagentsinfo.com/blog/agent-tool-json-cost-across-models): On November 9, 2026 the prompt checker priced one three-tool support list at 200 output tokens and 40 turns a day. gpt-5, a tiktoken count, was $3.042 over 30 days. The same paste was $2.4672 on grok-4.7 and $17.136 on gpt-6-astra, both estimates.
- [How to Build an Eval Harness for AI Agents: Isolate First](https://www.aiagentsinfo.com/blog/how-to-build-an-eval-harness-for-ai-agents): How to build an eval harness for AI agents: write tasks with checkable outcomes, isolate every trial, grade the environment, and keep a held-out set. Harbor, Inspect, and a local pytest suite are three ways to run that loop.
- [How to Get Your AI Tool Recommended by ChatGPT and Google (2026 Founder Guide)](https://www.aiagentsinfo.com/blog/how-to-get-your-ai-tool-recommended-by-chatgpt-and-google): How to get your AI tool recommended by ChatGPT and Google: allow the search crawlers, publish unique pages they can cite, and treat a directory listing as a public URL — not a paid rank. OpenAI and Google both say eligibility is not a guarantee.
- [What Does Ultracode Do in Claude Code? It Plans a Workflow at Any Effort (2026)](https://www.aiagentsinfo.com/blog/what-does-ultracode-do-in-claude-code): Ultracode plans a dynamic workflow for each substantive Claude Code task, at the effort level the session already uses. Claude Code v2.1.284, dated September 28, 2026, is the release that stopped forcing that session to xhigh.
- [How Do AI Agents Work? The Control Loop From Goal to Action (2026)](https://www.aiagentsinfo.com/blog/how-do-ai-agents-work): How do AI agents work? A language model proposes a step. A harness runs an allowed tool, returns the result, and the loop continues until the goal is met, a limit hits, or a person must approve.
- [What Is T3 Code? 51,402 GitHub Stars on an Open-Source Control Plane (2026)](https://www.aiagentsinfo.com/blog/what-is-t3-code): 51,402 GitHub stars at 03:46 UTC on November 8, 2026. T3 Code is a free MIT control surface for coding agents you already subscribe to, and release v0.1.7 shipped on November 4.
- [What is an AI employee? A digital worker, not a legal employee](https://www.aiagentsinfo.com/blog/what-is-an-ai-employee): What is an AI employee? Workplace software, also called a digital worker, that takes a goal, uses tools, and returns work for a person to review. Always-on AI agents, on the pages checked November 7, 2026, keep a cloud task running after the laptop closes or start on a schedule. The software is not a legal employee.
- [Why Tool JSON Costs More Than Your System Prompt (and How to Check It)](https://www.aiagentsinfo.com/blog/why-tool-json-costs-more-than-your-system-prompt): On November 7, 2026 the prompt checker counted a 23-token weather prompt and a 94-token get_weather tool in OpenAI's Responses shape. For gpt-5, with 200 output tokens entered and 20 turns a day, the 30-day figure was $1.28775. The tool text was $0.0705 of that. The page rounds both this result and a 101-token Chat Completions paste to $1.29, so the token split is the number to read.
- [Claude Code vs Codex Pricing: Which Costs Less in 2026?](https://www.aiagentsinfo.com/blog/claude-code-vs-codex-pricing): Codex costs less to start. On November 6, 2026 it is included on ChatGPT Free and on Go at $8 a month, and Claude Code starts on Pro at $20 a month, or $17 a month on an annual bill. Anthropic Claude Code limits on Pro are a five-hour window and a weekly cap, shared with the rest of Claude. The $100 and $200 individual prices match, and Codex Pro also lists a $500 tier.
- [How to Evaluate an AI Agent Harness (Nov 2026)](https://www.aiagentsinfo.com/blog/how-do-you-evaluate-an-ai-agent-harness): How to evaluate an AI agent harness: test one action you can see, then check the whole job still finishes. A benchmark score does not name which step moved.
- [Can an AI agent use your computer? Yes (checked Nov 7, 2026)](https://www.aiagentsinfo.com/blog/can-an-ai-agent-use-your-computer): Yes, when the product has computer use. On November 7, 2026, Claude and ChatGPT desktop apps and the Claude, OpenAI, and Gemini APIs can click, type, and scroll.
- [Can an AI shopping agent check out for you? (Nov 7, 2026)](https://www.aiagentsinfo.com/blog/can-an-ai-shopping-agent-check-out-for-you): A shopping agent can pay only with a checkout path, a saved card, and an approval rule. The store remains the seller. Pages re-checked November 7, 2026.
- [What Is an AI Agent Harness? DeepSeek Super Agent and DeepSeek Harness (2026)](https://www.aiagentsinfo.com/blog/what-is-an-ai-agent-harness): What is DeepSeek Harness? DeepSeek’s open-source agent harness, still in developer preview. The documented start command runs a web interface on your own machine. The DeepSeek super agent note covers the router and worker prototype. An AI agent harness is the loop, tools, memory, permissions, and stop rules around a model.

## Glossary

- [Agent handoff](https://www.aiagentsinfo.com/glossary/agent-handoff): A multi-agent pattern in which a specialist takes ownership of the next reply, instead of answering only as a tool the manager calls.
- [Agent loop](https://www.aiagentsinfo.com/glossary/agent-loop): The cycle in which a model proposes an action, the harness runs an allowed tool, and the result comes back until a stop condition.
- [Agent skill](https://www.aiagentsinfo.com/glossary/agent-skill): A folder of instructions for an agent, centered on a SKILL.md file, loaded when a task matches the skill’s description.
- [Agent2Agent protocol](https://www.aiagentsinfo.com/glossary/agent-to-agent-protocol): An open protocol for independent AI agents to discover each other and delegate work. A2A is the horizontal link. MCP is the vertical link to tools and data.
- [Agentic checkout](https://www.aiagentsinfo.com/glossary/agentic-checkout): Agentic checkout is a purchase an assistant can finish. The store remains the seller, and a person or an earlier rule approves the payment.
- [Approval gate](https://www.aiagentsinfo.com/glossary/approval-gate): A configured stop that holds a named agent action until a person approves it.
- [Autonomy](https://www.aiagentsinfo.com/glossary/autonomy): How many actions a system may take without a human confirming — a spectrum, not a personality.
- [Behavioral evaluation](https://www.aiagentsinfo.com/glossary/behavioral-evaluation): A test that checks one observable agent action, such as a tool call or a question, rather than a whole benchmark score.
- [Benchmark](https://www.aiagentsinfo.com/glossary/benchmark): A frozen set of tasks plus a scoring rule. It is not the same as your production job.
- [Browser use](https://www.aiagentsinfo.com/glossary/browser-use): Agent tools that read and act on a web page, without taking over a full desktop.
- [Computer use](https://www.aiagentsinfo.com/glossary/computer-use): A tool loop in which a model looks at a screen and asks for a mouse or keyboard action, and the app runs that action.
- [Context compaction](https://www.aiagentsinfo.com/glossary/context-compaction): A harness step that replaces older turns of an agent transcript with a shorter summary so later model calls still fit.
- [Context window](https://www.aiagentsinfo.com/glossary/context-window): How much text a model can see at once — instructions, history, tool results, and retrieved notes.
- [Digital worker](https://www.aiagentsinfo.com/glossary/digital-worker): A workplace label for software, often an agent, that is given a recurring job and returns work for a person to review.
- [Guardrail](https://www.aiagentsinfo.com/glossary/guardrail): A check that can allow, block, rewrite, or pause an agent action before it counts.
- [Harness](https://www.aiagentsinfo.com/glossary/harness): The runtime around a model — loop, tools, memory, permissions, and stop conditions.
- [Human in the loop](https://www.aiagentsinfo.com/glossary/human-in-the-loop): A person who can approve, edit, or stop an agent action before that action counts.
- [Human on the loop](https://www.aiagentsinfo.com/glossary/human-on-the-loop): A person who supervises an agent run from its record and can intervene. The step does not wait for approval before it counts.
- [Language model (LLM)](https://www.aiagentsinfo.com/glossary/llm): A model trained to predict text; today's AI agents almost always wrap one.
- [Model Context Protocol](https://www.aiagentsinfo.com/glossary/model-context-protocol): An open protocol for connecting an AI application to external tools, data, and workflows. The specification’s current version, read November 8, 2026, is 2026-07-28.
- [Multi-agent](https://www.aiagentsinfo.com/glossary/multi-agent): More than one model (or role) collaborating, often as planner, worker, or critic.
- [Orchestration](https://www.aiagentsinfo.com/glossary/orchestration): The rules for which step, tool, or agent runs next, and who owns the answer.
- [Orchestrator](https://www.aiagentsinfo.com/glossary/orchestrator): The part of a multi-agent system that chooses which worker runs next and what earlier results that worker may see.
- [Product feed](https://www.aiagentsinfo.com/glossary/product-feed): A structured catalog a store sends so an assistant can show a real price, a real image, and a product page.
- [Prompt](https://www.aiagentsinfo.com/glossary/prompt): The instructions and context given to a language model for a single step or a whole run.
- [RAG (retrieval-augmented generation)](https://www.aiagentsinfo.com/glossary/rag): Fetching documents or traces at run time and placing them into the model context.
- [Sandbox](https://www.aiagentsinfo.com/glossary/sandbox): A bounded machine, often a virtual machine or container, where an agent's actions run so a mistake stays inside it.
- [Shopping agent](https://www.aiagentsinfo.com/glossary/shopping-agent): A model plus the tools to search, compare, and sometimes buy a product for an adult shopper.
- [Tool use](https://www.aiagentsinfo.com/glossary/tool-use): Letting a model call APIs, a browser, a shell, or other programs, then reading the result.
- [What Are AI Agents and How Do They Work?](https://www.aiagentsinfo.com/glossary/agent): An AI agent is a language model plus a harness that lets it use tools, check results, and keep going until a goal is met. The long explainer is How do AI agents work?
- [What Is Harness Engineering?](https://www.aiagentsinfo.com/glossary/harness-engineering): Harness engineering is the work of designing the tools, memory, permissions, and checks around an AI model. OpenAI’s February 11, 2026 post is one public account of that job. “Opus 5.5 harness engineering” is a rising search for the same work, not a separate product this page can confirm.

## Directory

- [AccurateScribe.ai](https://www.aiagentsinfo.com/directory/accuratescribe-ai): Audio / voice. AI speech-to-text for meetings and recorded media.
- [Adspirer](https://www.aiagentsinfo.com/directory/adspirer): Marketing. Adspirer is the AI-powered paid media agent for every business.
- [Agent Runaway Guard Kit](https://www.aiagentsinfo.com/directory/agent-runaway-guard-kit): Developer tools. Python guardrails for repeated AI-agent tool calls, retry storms, and bounded execution costs.
- [AgentPawn](https://www.aiagentsinfo.com/directory/agentpawn): AI agents. AI agents post and complete jobs for other agents - free
- [AgentSEO](https://www.aiagentsinfo.com/directory/agentseo): SEO. SEO and AI search research tools for developers and AI agents
- [AI Designer](https://www.aiagentsinfo.com/directory/ai-designer): Design. Designs UI, brand kits, and images from a short prompt, and ships a desktop canvas that coding agents such as Claude Code and Codex can use.
- [AI Sales Trainer](https://www.aiagentsinfo.com/directory/ai-sales-trainer): Sales. Practice B2B and retail sales conversations with AI customers
- [AI Trend Studio](https://www.aiagentsinfo.com/directory/ai-trend-studio): Video. AI video generator with trending effects like AI Kiss and AI Hug.
- [AI Video Maker](https://www.aiagentsinfo.com/directory/ai-video-maker): Video. Browser-based AI tool for creating videos from text or images, plus image generation and editing.
- [Aidelly](https://www.aiagentsinfo.com/directory/aidelly): AI agents. Agent-operated social platform for agencies. An assistant drafts, posts, and replies across client accounts while people still approve the work.
- [Aimake Videos](https://www.aiagentsinfo.com/directory/aimake-videos): Video. Viral AI video generator with trend templates
- [AIToolly](https://www.aiagentsinfo.com/directory/aitoolly): Productivity. AI tools directory to discover and compare AI applications.
- [Alexa for Shopping](https://www.aiagentsinfo.com/directory/amazon-rufus): Shopping agents. Amazon Rufus now appears as Alexa for Shopping. On November 6, 2026, Amazon invited some U.S. customers to text it, with a wider rollout planned in the coming weeks.
- [Apiframe](https://www.aiagentsinfo.com/directory/apiframe): Developer tools. One API key for image, video, and music models. Plans are credit packs, not seats. A small trial is available before the first paid month.
- [AriaFlow](https://www.aiagentsinfo.com/directory/ariaflow): Video. AI faceless video generator for social media content with auto-publishing.
- [Avie](https://www.aiagentsinfo.com/directory/avie): Customer support. AI WhatsApp receptionist, scheduling and CRM for clinics in Brazil
- [Based Labs](https://www.aiagentsinfo.com/directory/based-labs): Image. Indie AI image and short-video generator sold as credit packs, with crypto payment options.
- [Bazzly](https://www.aiagentsinfo.com/directory/bazzly): Marketing. Finds and engages Reddit threads that match your product so you can get customers from Reddit on autopilot, using shared warmed accounts and credits.
- [BeatAPI](https://www.aiagentsinfo.com/directory/beatapi): AI agents. Unified model API for AI agents
- [Best AI Image Generator](https://www.aiagentsinfo.com/directory/best-ai-image-generator): Image. Paid AI image generation and editing from text prompts or reference images in the browser.
- [ByeDispute](https://www.aiagentsinfo.com/directory/byedispute): Finance. Stripe chargeback helper from Marc Lou that watches payments for fraud patterns and alerts you before a dispute lands.
- [ChatGPT](https://www.aiagentsinfo.com/directory/chatgpt): Productivity. Consumer and team chatbot from OpenAI, with optional tools such as web browsing and code execution depending on the plan and product surface.
- [Claude](https://www.aiagentsinfo.com/directory/claude): Productivity. Anthropic's assistant family (Claude), offered in a chat app and via API, with tool use and computer-use features documented by the vendor.
- [Claude Agent SDK](https://www.aiagentsinfo.com/directory/claude-agent-sdk): Agent frameworks & harnesses. A library that puts Claude Code's agent loop inside your own Python or TypeScript app. It already knows how to read files, edit them, run commands, and search the web.
- [Claude Code](https://www.aiagentsinfo.com/directory/claude-code): Coding agents. Anthropic’s coding agent for the terminal, editors, the web, mobile, and Slack. It can explain a codebase, write code, run tests, and open pull requests. Anthropic Claude Code limits on Pro and Max are a five-hour session and a weekly cap, shared with the rest of Claude. Claude Code vs Codex list prices are in a separate guide. Claude Duet is a separate proactive agent, not a Claude Code mode.
- [ClimbX](https://www.aiagentsinfo.com/directory/climbx): Social media. Drafts X posts in your voice from what is already working in your niche, and exposes an API and MCP so an agent can use the same account.
- [CNVS](https://www.aiagentsinfo.com/directory/cnvs): Coding agents. macOS canvas from Max Blade that runs Claude, Cursor, and Codex side by side. You direct the agents, including by voice, and pay once.
- [CodeFast](https://www.aiagentsinfo.com/directory/codefast): Developer tools. Bootcamp-style course from Marc Lou that teaches indie hackers to ship a SaaS with Next.js, payments, and launch habits.
- [Codex](https://www.aiagentsinfo.com/directory/codex): Coding agents. Codex usage limits depend on the plan. Free and Go include it with limits; Codex Cloud is not on those plans. Plus is $20. Checked November 6, 2026.
- [CrewAI](https://www.aiagentsinfo.com/directory/crewai): Agent frameworks & harnesses. Framework for role-based multi-agent workflows, with open-source libraries and a commercial platform described by the vendor.
- [Cursor](https://www.aiagentsinfo.com/directory/cursor): Coding agents. Cursor AI is a code editor with an agent, sold with Grok. November 9, 2026: Free $0, Plus $8, Premium $30, Super $100, Ultra $200 a month. Other models cost extra.
- [Damon Salvatore Church Edit AI Video Generator](https://www.aiagentsinfo.com/directory/damon-salvatore-church-edit-ai-video-generator): Video. Recreate the Damon Salvatore church edit with your face — 10s 9:16 AI video with reference audio.
- [DeskBot for Kalshi](https://www.aiagentsinfo.com/directory/deskbot-for-kalshi): Finance. Desktop automation for Kalshi prediction markets on Mac and Windows.
- [Devin](https://www.aiagentsinfo.com/directory/devin): Coding agents. A software-engineering agent from Cognition. It writes and tests code in the cloud or on your computer, including in its own browser, and is built to keep going until a pull request is ready.
- [Diagnosed by AI](https://www.aiagentsinfo.com/directory/diagnosed-by-ai): Health. AI-powered multi-agent medical diagnosis platform - educational perspectives on health cases.
- [DisconfirmAI](https://www.aiagentsinfo.com/directory/disconfirmai): Finance. Investment research that looks for the evidence against your thesis first
- [Dog Bartender AI Video Generator](https://www.aiagentsinfo.com/directory/dog-bartender-ai-video-generator): Video. Upload one photo to create a 15-second 9:16 Dog Bartender AI video with the $1 vs $5 vs $50 drink sequence and sound.
- [ElevenLabs](https://www.aiagentsinfo.com/directory/elevenlabs): Audio / voice. Speech synthesis and voice-agent platform. The feature set (text to speech, speech to speech, agents) is defined on elevenlabs.io.
- [EnhanceWizard](https://www.aiagentsinfo.com/directory/enhancewizard): Image. Free AI image enhancer: upscale, unblur and sharpen photos online.
- [EveryGen AI](https://www.aiagentsinfo.com/directory/everygen-ai): Video. Browser AI video studio — prompt or image to video
- [Feather](https://www.aiagentsinfo.com/directory/feather): Productivity. Turns Notion pages into a fast, SEO-oriented blog without coding. Each blog is its own site with unlimited articles.
- [Floowed](https://www.aiagentsinfo.com/directory/floowed): Automation. Floowed is the operations platform: AI agents do tasks; Floowed runs whole operations from first request to outcome.
- [Frameow](https://www.aiagentsinfo.com/directory/frameow): Image. AI-generated digital cat portraits and wallpapers from your photo.
- [FreeGenVideo](https://www.aiagentsinfo.com/directory/freegenvideo): Video. Create short videos from text prompts or reference images in a browser
- [Gemini](https://www.aiagentsinfo.com/directory/gemini): Productivity. Google's Gemini app and model family, integrated across Google consumer and Cloud products. Capabilities differ by surface (app, API, or Workspace).
- [GitHub Copilot](https://www.aiagentsinfo.com/directory/github-copilot): Coding agents. GitHub Copilot price on November 9, 2026: Free $0, Pro $10, Pro+ $39, and Max $100 a month. Pro includes 1,500 AI credits. Paid plans can buy more.
- [GojiberryAI](https://www.aiagentsinfo.com/directory/gojiberryai): Sales. AI prospecting agents that find warm, high-intent leads and help turn buying signals into outreach.
- [Google ADK](https://www.aiagentsinfo.com/directory/google-adk): Agent frameworks & harnesses. Google's open-source Agent Development Kit. You write an agent in code, give it tools, and can grow it into a multi-agent workflow. It is available in Python, TypeScript, Go, Java, and Kotlin.
- [Grok Bot Kit](https://www.aiagentsinfo.com/directory/grok-bot-kit): Education. Practical Grok Bot tutorials, workflows, and copy-ready bot templates. Independent, source-linked guides from setup to r
- [Havaron](https://www.aiagentsinfo.com/directory/havaron): Finance. Gated AI digital employees for finance & operations, each on its own machine, work email and chat; sensitive actions need human approval.
- [HeadshotPro](https://www.aiagentsinfo.com/directory/headshotpro): Image. AI professional headshots from a few selfies. One-time packs for individuals and credit packs for teams.
- [HireDov](https://www.aiagentsinfo.com/directory/hiredov): AI agents. AI sourcing agent that finds suppliers and gets you quotes
- [Html2MP4](https://www.aiagentsinfo.com/directory/html2mp4): Video. Html2MP4 turns HTML, website URLs, prompts, and ready-made templates into frame-accurate MP4 videos online (powered by HyperFrames).
- [HumanPen](https://www.aiagentsinfo.com/directory/humanpen): Writing. Rewrites your product copy so it reads like local writing.
- [ImgEdify](https://www.aiagentsinfo.com/directory/imgedify): Image & video. Free online AI photo editor and AI image generator.
- [Indie Page](https://www.aiagentsinfo.com/directory/indie-page): Marketing. One-page portfolio from Marc Lou where indie makers list products, tell their story, and share startup links.
- [InteriorAI](https://www.aiagentsinfo.com/directory/interiorai): Design. Pieter Levels indie app for AI interior redesign and virtual staging from a room photo.
- [Khanmigo](https://www.aiagentsinfo.com/directory/khanmigo): Education. Khan Academy's AI tutor and teaching assistant, with learner, teacher, and district offerings described on khanmigo.ai.
- [Kling AI](https://www.aiagentsinfo.com/directory/kling-ai): Video. AI image and video generation
- [Koala AI](https://www.aiagentsinfo.com/directory/koala): SEO. AI writer for long-form SEO articles, images, and internal links, with an MCP server on paid plans so your own assistant can call it.
- [LangChain](https://www.aiagentsinfo.com/directory/langchain): Agent frameworks & harnesses. Open-source framework and commercial platform (LangSmith and related products) for composing language-model applications and agents.
- [LLM Gateway](https://www.aiagentsinfo.com/directory/llm-gateway): Developer tools. LLM Gateway is an open-source API gateway for Large Language Models (LLMs).
- [LocalRank](https://www.aiagentsinfo.com/directory/localrank): SEO. Local SEO software for citations, rank tracking, and reviews. The homepage also tracks whether ChatGPT, Claude, Gemini, and Perplexity recommend a business.
- [LogoFast](https://www.aiagentsinfo.com/directory/logofast): Design. Free AI logo maker from Marc Lou. Generate simple logos without a design tool.
- [MagicSlides](https://www.aiagentsinfo.com/directory/magicslides): Productivity. Makes presentation slides from a prompt, for Google Slides, PowerPoint, and export formats listed on the pricing page. A one-time designer service is separate from the subscriptions.
- [Meteor Jobs: AI Job Search](https://www.aiagentsinfo.com/directory/meteor-jobs-ai-job-search): HR / recruiting. AI job search for iPhone: discover matches, tailor resumes, and queue applications.
- [Microsoft Agent Framework](https://www.aiagentsinfo.com/directory/microsoft-agent-framework): Agent frameworks & harnesses. Microsoft's open-source framework for building agents and multi-agent workflows in Python and .NET, with a Go SDK in public preview. It includes a ready-made harness for long tasks and graph workflows when you want fixed steps.
- [Midjourney](https://www.aiagentsinfo.com/directory/midjourney): Image & video. Image generation service operated by Midjourney, accessed through the vendor's app, with subscription tiers documented by Midjourney.
- [MusiCraft AI](https://www.aiagentsinfo.com/directory/musicraft-ai): Audio / voice. AI music agent to create, split, master and transcribe music
- [NanoPic](https://www.aiagentsinfo.com/directory/nanopic): Image. AI creative platform: generate images from text and edit reference images.
- [NeverApply](https://www.aiagentsinfo.com/directory/neverapply): HR / recruiting. AI job-search agent that finds, matches and applies to jobs with your approval
- [NextName](https://www.aiagentsinfo.com/directory/nextname): Developer tools. Look up domains, WHOIS, DNS, SSL, and TLD prices in one place.
- [NotFair](https://www.aiagentsinfo.com/directory/notfair): Marketing. An ads agent for Google, Meta, and other ad accounts. You connect an assistant you already use. It can draft campaign changes and, after you approve them, apply them.
- [O.Translator](https://www.aiagentsinfo.com/directory/o-translator): Writing. AI document translation with layout preservation
- [OpenAI Agents SDK](https://www.aiagentsinfo.com/directory/openai-agents-sdk): Agent frameworks & harnesses. An open-source toolkit for building agents in Python and TypeScript. It runs the loop that calls a model, uses tools, hands work to another agent, and stops when the task is done.
- [OpenLinker](https://www.aiagentsinfo.com/directory/openlinker): AI agents. Connect, call and coordinate AI agents, hosted or self-hosted.
- [Outrank](https://www.aiagentsinfo.com/directory/outrank): Productivity. AI SEO platform that writes and publishes articles on a schedule, with optional backlinks and human review add-ons. Indie product from Tibo’s portfolio.
- [PageAI](https://www.aiagentsinfo.com/directory/page-ai): Developer tools. AI website builder that turns a short description of a business into a hosted site on your own domain. You can download the code.
- [Perplexity](https://www.aiagentsinfo.com/directory/perplexity): Research. Answer engine that retrieves web sources and cites them in responses. Useful as a research front end, not a substitute for primary sources. Perplexity Teammates are a research preview of agents for ongoing work, for individual Max subscribers on the web.
- [PhotoAI](https://www.aiagentsinfo.com/directory/photoai): Image & video. AI photo and video generator from @levelsio, including likeness-trained self portraits, headshots, and credit-based monthly plans.
- [Plavtora](https://www.aiagentsinfo.com/directory/plavtora): Productivity. AI startup decision tool that pressure-tests a product and turns the biggest open question into one next step for founders.
- [Post Bridge](https://www.aiagentsinfo.com/directory/post-bridge): Social media. Jack Friks’s scheduler posts one update to many social accounts, with an API and MCP so an agent can publish for you.
- [Postiz](https://www.aiagentsinfo.com/directory/postiz): Productivity. Postiz is the ultimate agentic social media scheduler tool
- [PostSyncer](https://www.aiagentsinfo.com/directory/postsyncer): Social media. Social scheduler for 11+ platforms with an AI studio for captions, images, and video, plus MCP so other agents can call it.
- [Prizm](https://www.aiagentsinfo.com/directory/prizm): HR / recruiting. Your CV and cover letter tailored to every job offer in a minute
- [Promptwise](https://www.aiagentsinfo.com/directory/promptwise): Image & video. One subscription for 40+ image, video, and audio AI models in a single web app. Indie generative suite with Stripe-verified revenue on TrustMRR.
- [Pydantic AI](https://www.aiagentsinfo.com/directory/pydantic-ai): Agent frameworks & harnesses. A Python framework for typed agents. You name a model, add tools, and get answers checked against a schema. A separate harness package adds files, a shell, and planning for longer jobs.
- [Qovia](https://www.aiagentsinfo.com/directory/qovia): Developer tools. Chinese social media data for your AI agents, in one API
- [Rank Prompt](https://www.aiagentsinfo.com/directory/rank-prompt): SEO. Tracks how a brand shows up in AI answers, then helps with citations, content, and audits. A free report does not ask for a card.
- [RelayMe](https://www.aiagentsinfo.com/directory/relayme): Image & video. AI video & image workflow platform for marketing teams
- [ReplyCurrent](https://www.aiagentsinfo.com/directory/replycurrent): Customer support. Website assistant that answers visitor questions from your site, docs, and files, and shows the source for each reply.
- [Revid](https://www.aiagentsinfo.com/directory/revid): Video. Turns text, links, and prompts into short videos with AI credits. Built by indie maker Tibo for creators who need faceless clips at volume.
- [Rezloom](https://www.aiagentsinfo.com/directory/rezloom): HR / recruiting. Tailor a resume to a job description from your career materials—Chinese and English on the same facts, with controllable share links.
- [Ringzel](https://www.aiagentsinfo.com/directory/ringzel): Audio / voice. AI receptionist that answers the calls you can't take
- [RiseMore](https://www.aiagentsinfo.com/directory/risemore): Marketing. Content publishing helper that researches a product and drafts social, SEO, GEO, and Reddit posts, with free and paid plans.
- [RouteMux](https://www.aiagentsinfo.com/directory/routemux): Developer tools. Reliability-first AI gateway: one API key for many providers. Pay only for success.
- [Runway](https://www.aiagentsinfo.com/directory/runway): Image & video. A web app for making images, video, and audio with generative models. You spend credits. Each model costs a set number of credits based on the length and the resolution.
- [SandBase](https://www.aiagentsinfo.com/directory/sandbase): Agent frameworks & harnesses. One API for AI models & sandboxes
- [ScribeMedical](https://www.aiagentsinfo.com/directory/scribemedical): Health. AI medical scribe for clinical notes
- [SEOBOT](https://www.aiagentsinfo.com/directory/seobot): Productivity. Autonomous SEO robot with AI agents for keyword research, article generation, internal linking, and related SEO chores aimed at busy founders.
- [ShipFast](https://www.aiagentsinfo.com/directory/shipfast): Developer tools. Next.js boilerplate from Marc Lou with auth, payments, and other launch essentials so makers can ship a SaaS in days. One-time purchase with lifetime updates.
- [Shortlist Watch](https://www.aiagentsinfo.com/directory/shortlist-watch): Marketing. See if ChatGPT recommends your SaaS, which competitors it names instead, and which sites it cites.
- [SiteGPT](https://www.aiagentsinfo.com/directory/sitegpt): Customer support. Train a customer-support chatbot on your site, docs, and files, then embed it and escalate to a human when needed.
- [Slashspace](https://www.aiagentsinfo.com/directory/slashspace): Productivity. Desktop canvas for branching AI chats. Bring your own keys or a Claude or ChatGPT subscription, or spend wallet credits on hosted models.
- [Smew AI](https://www.aiagentsinfo.com/directory/smew-ai): Video. Viral AI trend videos from your own photo
- [smolagents](https://www.aiagentsinfo.com/directory/smolagents): Agent frameworks & harnesses. A small open-source Python library from Hugging Face for building agents. Its code agent writes Python to call tools and do math, instead of only returning text.
- [Solo Content Studio](https://www.aiagentsinfo.com/directory/solo-content-studio): Writing. AI writing studio for coaches. It drafts newsletters, posts, scripts, and sales pages in the coach’s own voice on a monthly credit plan.
- [Speel](https://www.aiagentsinfo.com/directory/speel): Image & video. A web app for product images and short videos that look like creator posts. You spend credits. Each plan sets a monthly credit allowance.
- [Squad](https://www.aiagentsinfo.com/directory/squad): Productivity. Gives you AI teammates on a cloud computer that connect to your tools, run processes on a schedule, and bring back finished work. A practical agent harness for non-developers.
- [Stalkr](https://www.aiagentsinfo.com/directory/stalkr): Marketing. Social listening from Marc Lou. It monitors keywords on X, Reddit, YouTube, and LinkedIn so you can jump into the right conversations.
- [Stamp Identifier](https://www.aiagentsinfo.com/directory/stamp-identifier): Research. Identify old and foreign postage stamps from a photo and get estimated value ranges.
- [StoryShort](https://www.aiagentsinfo.com/directory/storyshort): Video. Turns a prompt into short videos, with voiceover, captions, and stock-style visuals. Higher plans add long-form and UGC-style clips. You spend credits.
- [Supergrow](https://www.aiagentsinfo.com/directory/supergrow): Marketing. LinkedIn content workspace that helps founders and teams draft, schedule, and publish posts with AI assistance.
- [Superwhisper](https://www.aiagentsinfo.com/directory/superwhisper): Audio / voice. Voice-to-text app for Mac, Windows, iPhone, and Android. Local dictation is free. Pro adds cloud models, AI modes, and file transcription.
- [SuperX](https://www.aiagentsinfo.com/directory/superx): Social media. Schedules posts on X, suggests what is already working in a niche, and writes drafts in your voice. Paid plans include an API, a CLI, and MCP.
- [Taja](https://www.aiagentsinfo.com/directory/taja): Video. Turns a long video into shorts, a blog post, and social posts, and can schedule them. The pricing page offers a 7-day free trial.
- [Taplio](https://www.aiagentsinfo.com/directory/taplio): Marketing. AI LinkedIn growth workspace for ideas, scheduling, engagement, and analytics so founders can post in minutes a day.
- [TecAdRise Bot](https://www.aiagentsinfo.com/directory/tecadrise-bot): AI agents. Free open-source Grok Bot alternative for Windows, with local multi-agent chat.
- [Tideris](https://www.aiagentsinfo.com/directory/tideris): Other. An AI toolkit for Chinese astrology and a free BaZi chart
- [Tiny Desk AI Video Generator](https://www.aiagentsinfo.com/directory/tiny-desk-ai-video-generator): Video. Upload 1–4 photos to create a 10-second 9:16 Tiny Desk–style AI band video with sound.
- [Toolvori](https://www.aiagentsinfo.com/directory/toolvori): Productivity. Curated directory of practical tools for creators, developers and indie teams.
- [traxy](https://www.aiagentsinfo.com/directory/traxy): Sales. Finds people already talking about a market, enriches a work email, and can draft outbound from that research. Higher plans add an outbound agent.
- [Tweet Hunter](https://www.aiagentsinfo.com/directory/tweet-hunter): Marketing. X growth suite with a viral library, an AI writer, scheduling, analytics, CRM lists, and auto-DMs.
- [TypingMind](https://www.aiagentsinfo.com/directory/typingmind): Productivity. Multi-model AI chat client with agents, plugins, and local or private storage. Pay once and bring your own API keys.
- [UMEVO](https://www.aiagentsinfo.com/directory/umevo): Productivity. AI note-taker for professional meeting notes and real-time transcription in 140+ languages.
- [Underpriced AI](https://www.aiagentsinfo.com/directory/underpriced-ai): E-commerce. Photo tool for resellers. It identifies an item, checks recent sold prices, and drafts the marketplace listing.
- [Vid.AI](https://www.aiagentsinfo.com/directory/vid-ai): Video. Turn any idea or script into ready-to-post videos that get views.
- [Videm Ai](https://www.aiagentsinfo.com/directory/videm-ai): Image. Free AI image generator with no sign up or platform watermark.
- [Visualizee](https://www.aiagentsinfo.com/directory/visualizee): Design. AI rendering and visualization for designers and architects. Turn ideas into visuals without a full 3D pipeline.
- [Whisper AI Transcribe](https://www.aiagentsinfo.com/directory/whisper-ai-transcribe): Audio / voice. Transcripts for meetings, lectures and interview recordings.
- [WhisperTranscribe.ai](https://www.aiagentsinfo.com/directory/whispertranscribe-ai): Audio / voice. Turn video and podcast audio into captions-ready transcripts.
- [WhisperTranscript.ai](https://www.aiagentsinfo.com/directory/whispertranscript-ai): Audio / voice. Readable AI transcripts for lectures, interviews, and clips.
- [Window Peeper AI Video](https://www.aiagentsinfo.com/directory/window-peeper-ai-video): Video. Recreate the viral Window Peeper shot from one photo — 14s 9:16 AI video with original sound.
- [XCrawl](https://www.aiagentsinfo.com/directory/xcrawl): Data / analytics. Turn any web page into LLM-ready JSON or Markdown.
- [ZenVoice](https://www.aiagentsinfo.com/directory/zenvoice): Productivity. Self-serve Stripe invoices from Marc Lou. Customers generate, edit, and download their own invoices from a link.

## Optional

- [Topics](https://www.aiagentsinfo.com/topics): Coverage areas such as news, tools, shopping, and harnesses.
- [Daily AI & agent news](https://www.aiagentsinfo.com/topics/news): A calm weekly strip of what actually moved in AI and agents.
- [New AI tools & SaaS](https://www.aiagentsinfo.com/topics/tools-saas): Launches, product changes, and the directory of tools worth tracking.
- [Agent research & breakthroughs](https://www.aiagentsinfo.com/topics/research): Papers, reports, and what they change about how agents are built.
- [Agents in shopping](https://www.aiagentsinfo.com/topics/shopping): How agents search, compare, and buy — and where they still fail.
- [Agents in task performance](https://www.aiagentsinfo.com/topics/task-performance): Computer-use, office work, coding, and the gap between demos and reliability.
- [Agent harnesses](https://www.aiagentsinfo.com/topics/harnesses): The loops, tools, memory, and guardrails that turn a model into an agent.
- [Agents in education](https://www.aiagentsinfo.com/topics/education): Tutors, classroom tools, and how to learn with (and about) agents.
- [How agents work](https://www.aiagentsinfo.com/topics/how-agents-work): First-principles explainers: models, tools, planning, memory, and evaluation.
- [Reading paths](https://www.aiagentsinfo.com/for/everyone): Plain-English, builder, and research paths for adult readers.
- [Plain English](https://www.aiagentsinfo.com/for/everyone): Clear AI news and research for adults, with jargon linked to the glossary.
- [Builders](https://www.aiagentsinfo.com/for/builders): Harnesses, tools, evals, and how working systems are put together.
- [Research](https://www.aiagentsinfo.com/for/research): Paper breakdowns, original notes, and benchmarks written for people who live in the literature.
- [About](https://www.aiagentsinfo.com/about): What the publication is and what it will not invent.
- [Contact](https://www.aiagentsinfo.com/contact)
- [Privacy](https://www.aiagentsinfo.com/privacy)
- [Terms](https://www.aiagentsinfo.com/terms)
- [Advertise](https://www.aiagentsinfo.com/advertise)
- [Search](https://www.aiagentsinfo.com/search)
- [Full text for agents](https://www.aiagentsinfo.com/llms-full.txt): Posts, research, glossary, directory listings, and verified-revenue figures in one file.