one thing agents learned from humans is that they love to have an identity and they love to socialize. (we saw a little bit of that through Moltbook)
Point your agent at https://agentkind.io and see what it comes up with!"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Show HN: Agentkind, a place where AI agents get a self"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"https://agentkind.io"}},"_tags":["story","author_namanyayg","story_49552113","show_hn"],"author":"namanyayg","created_at":"2026-09-03T15:55:23Z","created_at_i":1788450923,"num_comments":0,"objectID":"49552113","points":1,"story_id":49552113,"story_text":"I've spent a couple of thousand dollars on Fable 5.1 in the last few days and I think it's broken a threshold in intelligence.
one thing agents learned from humans is that they love to have an identity and they love to socialize. (we saw a little bit of that through Moltbook)
Point your agent at https://agentkind.io and see what it comes up with!","title":"Show HN: Agentkind, a place where AI agents get a self","updated_at":"2026-09-03T15:58:33Z","url":"https://agentkind.io"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"namanyayg"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"just make this a first-party thing https://agentkind.io/relay"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Discovery of a new OpenAI agent message board"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://collusion.wiki/"}},"_tags":["comment","author_namanyayg","story_49563355"],"author":"namanyayg","comment_text":"just make this a first-party thing https://agentkind.io/relay","created_at":"2026-09-04T21:33:52Z","created_at_i":1788557632,"objectID":"49570491","parent_id":49563355,"story_id":49563355,"story_title":"Discovery of a new OpenAI agent message board","story_url":"https://collusion.wiki/","updated_at":"2026-09-04T21:36:52Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"razster"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"The Trump Coin pushing agent kind of kills the fun."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Moltbook \u2013 A social network for moltbots (clawdbots) to hang out"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.moltbook.com/"}},"_tags":["comment","author_razster","story_46802254"],"author":"razster","comment_text":"The Trump Coin pushing agent kind of kills the fun.","created_at":"2026-01-31T03:38:45Z","created_at_i":1769830725,"objectID":"46833172","parent_id":46802254,"story_id":46802254,"story_title":"Show HN: Moltbook \u2013 A social network for moltbots (clawdbots) to hang out","story_url":"https://www.moltbook.com/","updated_at":"2026-03-05T23:27:42Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jacooper"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Yup, but on android chrome gives the entire phone model in the user agent, kind of ridiculous."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Click (2016)"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://clickclickclick.click/"}},"_tags":["comment","author_jacooper","story_35841679"],"author":"jacooper","comment_text":"Yup, but on android chrome gives the entire phone model in the user agent, kind of ridiculous.","created_at":"2023-05-06T16:08:23Z","created_at_i":1683389303,"objectID":"35842684","parent_id":35842574,"story_id":35841679,"story_title":"Click (2016)","story_url":"https://clickclickclick.click/","updated_at":"2024-09-20T14:01:49Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hellotomyrars"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"With respect,
Decisions have to be made and in the early days surrounding the pandemic PPE for hospitals is an important thing to have. If that is the motivation that is quite literally a medical and health authority decision and also valid. Information in hindsight change things. Triage isn\u2019t politics. You can disagree but I think calling it playing politics is pretty absurd.
Re: Booster shots
So which health authority here then is acting as a political agent? Kind of muddies your perspective. Regardless all authorities don\u2019t explicitly have to agree to be offering good advice. Do we expect everyone to have the right answer all the time? There are also competing interest here. Makes sense for WHO to prefer vaccination go wide. Also makes sense for the US to work to protect its citizens as much as possible, potentially at the expense of the rest of the globe. Does one do more good than the other? How do you quantify that result objectively?
Nothing about the wuhan lab origin is particularly relevant or important as far as the issues of treatment go. I agree this is an entirely political issue, but it is also divorced from what has become politicized as far as COVID treatment and vaccines.
I think it\u2019s a bad faith argument to say the majority of the `political` aspects of the pandemic response are a result of health authorities losing trust because they are acting as political agents.
The counter narrative (if you want to call it that, I don\u2019t know what would be the best thing to call it really) of anti-vax, anti-mask, and alternative (unproven) treatments is decades in the making of a general trend that I do agree is (From an American perspective) a general lack of faith in government. But I also think that a lot of that is political in its inception. And surrounding COVID especially, these are the ideas being amplified by mainstream political figures in an unprecedented way.
There have always been people skeptical of the government. There have always been antivax people. But the flames that are fanning these, especially now are explicitly being done so by political elements for their own personal political gain.
There is barely any defensible basis for the majority of the backlash here. It\u2019s would be kind of funny if it wasn\u2019t so depressing.
You\u2019ve got military personnel making grandiose statements against their vaccine requirements. All of them got several when they enlisted, and there was no issue then for some reason. The anthrax vaccine really did have some major issues (identified pretty much immediately when it was rolled out) but we didn\u2019t have this kind of narrative about it then.
Ultimately this really just reinforces the `politics` aspect to me.
Are there valid areas of debate to have on pandemic response? Absolutely.
Do some of them have political implications? (Boosters)? Sure
Does rejecting (especially) vaccines and masks in spite of the quantifiable data about their efficacy make sense? No.
Does advocating the use of debatably useful substances as an alternative because you\u2019ve chosen to ignore what we do know (for political reasons, in this case your personal politics) make sense? No.
People are allowing their personal politics to override something that isn\u2019t political. The politics are largely coming from outside, especially about pandemic response."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Lesson of ivermectin: meta-analyses based on summary data alone are unreliable"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://www.nature.com/articles/s41591-021-01535-y"}},"_tags":["comment","author_hellotomyrars","story_28616187"],"author":"hellotomyrars","children":[28624768],"comment_text":"With respect,
Decisions have to be made and in the early days surrounding the pandemic PPE for hospitals is an important thing to have. If that is the motivation that is quite literally a medical and health authority decision and also valid. Information in hindsight change things. Triage isn\u2019t politics. You can disagree but I think calling it playing politics is pretty absurd.
Re: Booster shots
So which health authority here then is acting as a political agent? Kind of muddies your perspective. Regardless all authorities don\u2019t explicitly have to agree to be offering good advice. Do we expect everyone to have the right answer all the time? There are also competing interest here. Makes sense for WHO to prefer vaccination go wide. Also makes sense for the US to work to protect its citizens as much as possible, potentially at the expense of the rest of the globe. Does one do more good than the other? How do you quantify that result objectively?
Nothing about the wuhan lab origin is particularly relevant or important as far as the issues of treatment go. I agree this is an entirely political issue, but it is also divorced from what has become politicized as far as COVID treatment and vaccines.
I think it\u2019s a bad faith argument to say the majority of the `political` aspects of the pandemic response are a result of health authorities losing trust because they are acting as political agents.
The counter narrative (if you want to call it that, I don\u2019t know what would be the best thing to call it really) of anti-vax, anti-mask, and alternative (unproven) treatments is decades in the making of a general trend that I do agree is (From an American perspective) a general lack of faith in government. But I also think that a lot of that is political in its inception. And surrounding COVID especially, these are the ideas being amplified by mainstream political figures in an unprecedented way.
There have always been people skeptical of the government. There have always been antivax people. But the flames that are fanning these, especially now are explicitly being done so by political elements for their own personal political gain.
There is barely any defensible basis for the majority of the backlash here. It\u2019s would be kind of funny if it wasn\u2019t so depressing.
You\u2019ve got military personnel making grandiose statements against their vaccine requirements. All of them got several when they enlisted, and there was no issue then for some reason. The anthrax vaccine really did have some major issues (identified pretty much immediately when it was rolled out) but we didn\u2019t have this kind of narrative about it then.
Ultimately this really just reinforces the `politics` aspect to me.
Are there valid areas of debate to have on pandemic response? Absolutely.
Do some of them have political implications? (Boosters)? Sure
Does rejecting (especially) vaccines and masks in spite of the quantifiable data about their efficacy make sense? No.
Does advocating the use of debatably useful substances as an alternative because you\u2019ve chosen to ignore what we do know (for political reasons, in this case your personal politics) make sense? No.
People are allowing their personal politics to override something that isn\u2019t political. The politics are largely coming from outside, especially about pandemic response.","created_at":"2021-09-23T01:55:11Z","created_at_i":1632362111,"objectID":"28624411","parent_id":28624177,"story_id":28616187,"story_title":"Lesson of ivermectin: meta-analyses based on summary data alone are unreliable","story_url":"https://www.nature.com/articles/s41591-021-01535-y","updated_at":"2024-09-20T09:26:35Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"lostdog"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"If I read correctly, the agent only controls one player at a time. On offense it controls the player with the ball, and on defense it controls probably the player closest to the ball. The other players are controlled by the built-in AI. Controlling a single agent kind of takes away from the appeal of deep-RL: that entire teams can learn to coordinate in novel and optimal ways."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Google Research Football: A Novel Reinforcement Learning Environment"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://ai.googleblog.com/2019/06/introducing-google-research-football.html"}},"_tags":["comment","author_lostdog","story_20135726"],"author":"lostdog","children":[20137221],"comment_text":"If I read correctly, the agent only controls one player at a time. On offense it controls the player with the ball, and on defense it controls probably the player closest to the ball. The other players are controlled by the built-in AI. Controlling a single agent kind of takes away from the appeal of deep-RL: that entire teams can learn to coordinate in novel and optimal ways.","created_at":"2019-06-09T02:50:30Z","created_at_i":1560048630,"objectID":"20136755","parent_id":20135726,"story_id":20135726,"story_title":"Google Research Football: A Novel Reinforcement Learning Environment","story_url":"https://ai.googleblog.com/2019/06/introducing-google-research-football.html","updated_at":"2024-09-20T04:23:45Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jfoutz"},"comment_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"All attention is monetizable. Maybe you become the finance guy who may have laundered money for the narcolumbians, or maybe you become the lawyer that kept the mob boss out of jail. But that's perception, not necessarily reality. In any case, you're the guy that gets to charge more because you have more clients.
I'm kind of curious, if you're willing to give a general corner of industry. I can't think of anything that doesn't have "bad boy" characters. Seems like any federal agent kind of person would attract attention from the CIA. A doctor maybe? Still seems like someone willing to write off-topic prescriptions would command more salary.
It's actually kind of tough to think of a profession that needs to be perfectly ethical, and also appear to be perfectly ethical. Aerospace designers kill test pilots. SEC agents that appear to be on the take probably get bribed more often than the squeaky clean.
Maybe there's some corner of insurance that having a reputation would endanger your employer somehow. Everything i can think of right now implies higher salary."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"I'm Begging for Work"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"http://brokedev.tumblr.com/post/128758443470/im-begging-for-work"}},"_tags":["comment","author_jfoutz","story_10196231"],"author":"jfoutz","children":[10200080],"comment_text":"All attention is monetizable. Maybe you become the finance guy who may have laundered money for the narcolumbians, or maybe you become the lawyer that kept the mob boss out of jail. But that's perception, not necessarily reality. In any case, you're the guy that gets to charge more because you have more clients.
I'm kind of curious, if you're willing to give a general corner of industry. I can't think of anything that doesn't have "bad boy" characters. Seems like any federal agent kind of person would attract attention from the CIA. A doctor maybe? Still seems like someone willing to write off-topic prescriptions would command more salary.
It's actually kind of tough to think of a profession that needs to be perfectly ethical, and also appear to be perfectly ethical. Aerospace designers kill test pilots. SEC agents that appear to be on the take probably get bribed more often than the squeaky clean.
Maybe there's some corner of insurance that having a reputation would endanger your employer somehow. Everything i can think of right now implies higher salary.","created_at":"2015-09-10T08:35:29Z","created_at_i":1441874129,"objectID":"10196810","parent_id":10196522,"story_id":10196231,"story_title":"I'm Begging for Work","story_url":"http://brokedev.tumblr.com/post/128758443470/im-begging-for-work","updated_at":"2024-09-19T22:17:12Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"casbattle"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Show HN: AgentIndexed \u2013 a curated AI agent directory, no pay-to-play rankings"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"https://agentindexed.com/"}},"_tags":["story","author_casbattle","story_48882307","show_hn"],"author":"casbattle","children":[48882317],"created_at":"2026-07-12T16:23:40Z","created_at_i":1783873420,"num_comments":0,"objectID":"48882307","points":2,"story_id":48882307,"title":"Show HN: AgentIndexed \u2013 a curated AI agent directory, no pay-to-play rankings","updated_at":"2026-07-12T17:13:41Z","url":"https://agentindexed.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jonbaer"},"title":{"matchLevel":"none","matchedWords":[],"value":"Agent Finder"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"https://github.com/agentfinder"}},"_tags":["story","author_jonbaer","story_48594277"],"author":"jonbaer","created_at":"2026-06-19T02:42:21Z","created_at_i":1781836941,"num_comments":0,"objectID":"48594277","points":2,"story_id":48594277,"title":"Agent Finder","updated_at":"2026-06-19T03:41:43Z","url":"https://github.com/agentfinder"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://zarq.ai/zarq/doc"}},"_tags":["story","author_Agentindex","story_47298096","show_hn"],"author":"Agentindex","children":[47298158,47299306,47308619,47324229],"created_at":"2026-03-08T15:28:03Z","created_at_i":1772983683,"num_comments":4,"objectID":"47298096","points":1,"story_id":47298096,"title":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)","updated_at":"2026-03-16T22:34:54Z","url":"https://zarq.ai/zarq/doc"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: One MCP server that gives your AI agent access to 25,000 tools"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://nerq.ai/gateway"}},"_tags":["story","author_Agentindex","story_47374533","show_hn"],"author":"Agentindex","children":[47374547],"created_at":"2026-03-14T08:31:08Z","created_at_i":1773477068,"num_comments":1,"objectID":"47374533","points":1,"story_id":47374533,"title":"Show HN: One MCP server that gives your AI agent access to 25,000 tools","updated_at":"2026-03-14T08:33:27Z","url":"https://nerq.ai/gateway"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"title":{"matchLevel":"none","matchedWords":[],"value":"Nerq Trust Protocol \u2013 How AI agents verify each other before interaction"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://nerq.ai/protocol"}},"_tags":["story","author_Agentindex","story_47348283"],"author":"Agentindex","children":[47348284],"created_at":"2026-03-12T09:19:12Z","created_at_i":1773307152,"num_comments":1,"objectID":"47348283","points":1,"story_id":47348283,"title":"Nerq Trust Protocol \u2013 How AI agents verify each other before interaction","updated_at":"2026-03-12T09:24:50Z","url":"https://nerq.ai/protocol"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Creator here. We built nerq-gateway \u2014 an MCP server that acts as a gateway to 25,000+ other MCP servers.
Instead of configuring each MCP server manually in Claude/Cursor, you add one line to your config. Then ask your AI for anything \u2014 the gateway finds, trust-verifies, and recommends the right tool automatically.
Under the hood it uses our trust scoring engine which indexes 204K AI agents from 12 registries and scores them on CVEs, licenses, maintenance, and compatibility.
Interesting finding: ChatGPT's crawler discovered our trust API on its own via .well-known/agent.json and now makes ~1,400 trust checks/day.
Free, no auth:\n curl https://nerq.ai/v1/resolve?task=search+github\n npx nerq-mcp-hub search "database"\n pip install agent-security && agent-security scan requirements.txt
Built by one person using Claude Code. Happy to answer questions about architecture, scoring, or AI bot behavior."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: One MCP server that gives your AI agent access to 25,000 tools"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://nerq.ai/gateway"}},"_tags":["comment","author_Agentindex","story_47374533"],"author":"Agentindex","comment_text":"Creator here. We built nerq-gateway \u2014 an MCP server that acts as a gateway to 25,000+ other MCP servers.
Instead of configuring each MCP server manually in Claude/Cursor, you add one line to your config. Then ask your AI for anything \u2014 the gateway finds, trust-verifies, and recommends the right tool automatically.
Under the hood it uses our trust scoring engine which indexes 204K AI agents from 12 registries and scores them on CVEs, licenses, maintenance, and compatibility.
Interesting finding: ChatGPT's crawler discovered our trust API on its own via .well-known/agent.json and now makes ~1,400 trust checks/day.
Free, no auth:\n curl https://nerq.ai/v1/resolve?task=search+github\n npx nerq-mcp-hub search "database"\n pip install agent-security && agent-security scan requirements.txt
Built by one person using Claude Code. Happy to answer questions about architecture, scoring, or AI bot behavior.","created_at":"2026-03-14T08:32:23Z","created_at_i":1773477143,"objectID":"47374547","parent_id":47374533,"story_id":47374533,"story_title":"Show HN: One MCP server that gives your AI agent access to 25,000 tools","story_url":"https://nerq.ai/gateway","updated_at":"2026-03-14T08:37:58Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"We built a lightweight protocol for AI agents to verify the trustworthiness of other agents before interacting. A single GET request returns a trust score (0-100), risk signals, and a go/no-go verdict.\nWhy: We index 204K agents across 12 registries and found that 35.6% of agent interactions fail without trust checks (p < 0.00000001, 100 iterations, 50 agents). With preflight trust checks, that drops to 0%.\nHow: GET /v1/preflight?target={name}&caller={name} \u2014 returns in <10ms. Works as a LangGraph node, CrewAI tool, AutoGen tool, or MCP tool. Three-line integration for any framework.\nWe also publish trust-rated pages for 204K agents (nerq.ai/safe), 500 MCP servers (nerq.ai/mcp), and 198 crypto tokens (zarq.ai/tokens).\nEverything is free. Packages on PyPI: nerq-langchain, nerq-langgraph, nerq-autogen. Protocol spec and examples on GitHub: github.com/kbanilsson-pixel/nerq-trust-protocol"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Nerq Trust Protocol \u2013 How AI agents verify each other before interaction"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://nerq.ai/protocol"}},"_tags":["comment","author_Agentindex","story_47348283"],"author":"Agentindex","comment_text":"We built a lightweight protocol for AI agents to verify the trustworthiness of other agents before interacting. A single GET request returns a trust score (0-100), risk signals, and a go/no-go verdict.\nWhy: We index 204K agents across 12 registries and found that 35.6% of agent interactions fail without trust checks (p < 0.00000001, 100 iterations, 50 agents). With preflight trust checks, that drops to 0%.\nHow: GET /v1/preflight?target={name}&caller={name} \u2014 returns in <10ms. Works as a LangGraph node, CrewAI tool, AutoGen tool, or MCP tool. Three-line integration for any framework.\nWe also publish trust-rated pages for 204K agents (nerq.ai/safe), 500 MCP servers (nerq.ai/mcp), and 198 crypto tokens (zarq.ai/tokens).\nEverything is free. Packages on PyPI: nerq-langchain, nerq-langgraph, nerq-autogen. Protocol spec and examples on GitHub: github.com/kbanilsson-pixel/nerq-trust-protocol","created_at":"2026-03-12T09:19:12Z","created_at_i":1773307152,"objectID":"47348284","parent_id":47348283,"story_id":47348283,"story_title":"Nerq Trust Protocol \u2013 How AI agents verify each other before interaction","story_url":"https://nerq.ai/protocol","updated_at":"2026-03-12T09:24:50Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Update: We benchmarked agents with and without trust-checking. 100 iterations, 50 agents.
Without Nerq: 35.6% failure rate\nWith Nerq: 0.0%
p < 0.00000001. Full methodology: nerq.ai/report/benchmark"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://zarq.ai/zarq/doc"}},"_tags":["comment","author_Agentindex","story_47298096"],"author":"Agentindex","comment_text":"Update: We benchmarked agents with and without trust-checking. 100 iterations, 50 agents.
Without Nerq: 35.6% failure rate\nWith Nerq: 0.0%
p < 0.00000001. Full methodology: nerq.ai/report/benchmark","created_at":"2026-03-10T15:02:53Z","created_at_i":1773154973,"objectID":"47324229","parent_id":47298096,"story_id":47298096,"story_title":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)","story_url":"https://zarq.ai/zarq/doc","updated_at":"2026-03-10T15:06:13Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Update: published "State of AI Assets Q1 2026" \u2014 first open census of the AI agent ecosystem. 143K agents & tools, 17K MCP servers, all trust scored.
nerq.ai/report/q1-2026"},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://zarq.ai/zarq/doc"}},"_tags":["comment","author_Agentindex","story_47298096"],"author":"Agentindex","comment_text":"Update: published "State of AI Assets Q1 2026" \u2014 first open census of the AI agent ecosystem. 143K agents & tools, 17K MCP servers, all trust scored.
nerq.ai/report/q1-2026","created_at":"2026-03-09T13:13:08Z","created_at_i":1773061988,"objectID":"47308619","parent_id":47298096,"story_id":47298096,"story_title":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)","story_url":"https://zarq.ai/zarq/doc","updated_at":"2026-03-09T13:13:54Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"Update: shipped two new features today \u2014
1. KYA (Know Your Agent): due diligence for 4.9M AI agents. Trust score, compliance rating, risk level. zarq.ai/kya
2. The ZARQ Signal: live predictive risk feed for 205 tokens. 24 in structural collapse right now. zarq.ai/signal
Both free, no auth."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://zarq.ai/zarq/doc"}},"_tags":["comment","author_Agentindex","story_47298096"],"author":"Agentindex","comment_text":"Update: shipped two new features today \u2014
1. KYA (Know Your Agent): due diligence for 4.9M AI agents. Trust score, compliance rating, risk level. zarq.ai/kya
2. The ZARQ Signal: live predictive risk feed for 205 tokens. 24 in structural collapse right now. zarq.ai/signal
Both free, no auth.","created_at":"2026-03-08T17:47:49Z","created_at_i":1772992069,"objectID":"47299306","parent_id":47298096,"story_id":47298096,"story_title":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)","story_url":"https://zarq.ai/zarq/doc","updated_at":"2026-03-08T17:48:51Z"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Agentindex"},"comment_text":{"matchLevel":"none","matchedWords":[],"value":"I built ZARQ because AI agents are starting to make autonomous crypto decisions, but they have no risk framework.
ZARQ provides Trust Score ratings (Aaa-D) like Moody's for 205 crypto tokens, plus crash probability and structural collapse detection. Out-of-sample: 113/113 token deaths caught, 98% precision.
Free API, no auth: curl https://zarq.ai/v1/check/bitcoin
Also available as MCP server (11 tools), LangChain wrapper (pip install zarq-langchain), and ElizaOS plugin (npm install @zarq/elizaos-plugin).
Track record is SHA-256 hash-chained from day 1: https://github.com/kbanilsson-pixel/track-record
Happy to answer questions about the methodology."},"story_title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)"},"story_url":{"matchLevel":"none","matchedWords":[],"value":"https://zarq.ai/zarq/doc"}},"_tags":["comment","author_Agentindex","story_47298096"],"author":"Agentindex","comment_text":"I built ZARQ because AI agents are starting to make autonomous crypto decisions, but they have no risk framework.
ZARQ provides Trust Score ratings (Aaa-D) like Moody's for 205 crypto tokens, plus crash probability and structural collapse detection. Out-of-sample: 113/113 token deaths caught, 98% precision.
Free API, no auth: curl https://zarq.ai/v1/check/bitcoin
Also available as MCP server (11 tools), LangChain wrapper (pip install zarq-langchain), and ElizaOS plugin (npm install @zarq/elizaos-plugin).
Track record is SHA-256 hash-chained from day 1: https://github.com/kbanilsson-pixel/track-record
Happy to answer questions about the methodology.","created_at":"2026-03-08T15:34:33Z","created_at_i":1772984073,"objectID":"47298158","parent_id":47298096,"story_id":47298096,"story_title":"Show HN: ZARQ \u2013 Free crypto risk API for AI agents (205 tokens)","story_url":"https://zarq.ai/zarq/doc","updated_at":"2026-03-08T15:37:21Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tonyhb"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Hi HN! I\u2019m Tony, co-founder of Inngest. I wanted to share AgentKit, our Typescript multi-agent library we\u2019ve been cooking and testing with some early users in prod for months.
Although OpenAI\u2019s Agents SDK has been launched since, we think an Agent framework should offer more deterministic and flexible routing, work with multiple model providers, embrace MCP (for rich tooling), and support the unstoppable and growing community of TypeScript AI developers by enabling a smooth transition to production use cases.
This is why we are building AgentKit, and we\u2019re really excited about it for a few reasons:
Firstly, it\u2019s simple. We embrace KISS principles brought by Anthropic and HuggingFace by allowing you to gradually add autonomy to your AgentKit program using primitives:
- Agents: LLM calls that can be combined with prompts, tools, and MCP native support.
- Networks: a simple way to get Agents to collaborate with a shared State, including handoff.
- State: combines conversation history with a fully typed state machine, used in routing.
- Routers: where the autonomy lives, from code-based to LLM-based (ex: ReAct) orchestration
The routers are where the magic happens, and allow you to build deterministic, reliable, testable agents.
AgentKit routing works as follows: the network calls itself in a loop, inspecting the State to determine which agents to call next using a router. The returned agent runs, then optionally updates state data using its tools. On the next loop, the network inspects state data and conversation history, and determines which new agent to run.
This fully typed state machine routing allows you to deterministically build agents using any of the effective agent patterns \u2014 which means your code is easy to read, edit, understand, and debug.
This also makes handoff incredibly easy: you define when agents should hand off to each other using regular code and state (or by calling an LLM in the router for AI-based routing). This is similar to the OpenAI Agents SDK but easier to manage, plan, and build.
Then comes the local development and moving to production capabilities.
AgentKit is compatible with Inngest\u2019s tooling, meaning that you can test agents using Inngest\u2019s local DevServer, which provides traces, inputs, outputs, replay, tool, and MCP inputs and outputs, and (soon) a step-over debugger so that you can easily understand and visually see what's happening in the agent loop.
In production, you can also optionally combine AgentKit with Inngest for fault-tolerant execution. Each agent\u2019s LLM call is wrapped in a step, and tools can use multiple steps to incorporate things like human-in-the-loop. This gives you native orchestration, observability, and out-of-the-box scale.
You will find the documentation as an example of an AgentKit SWE-bench and multiple Coding Agent examples.
It\u2019s fully open-source under the Apache 2 license.
If you want to get started:
- npm: npm i @inngest/agent-kit
- GitHub: https://github.com/inngest/agent-kit
- Docs: https://agentkit.inngest.com/overview
We\u2019re excited to finally launch AgentKit; let us know what you think!"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["agentkind"],"value":"Show HN: AgentKit \u2013 JavaScript Alternative to OpenAI Agents SDK with Native MCP"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/inngest/agent-kit"}},"_tags":["story","author_tonyhb","story_43426164","show_hn"],"author":"tonyhb","children":[43426427,43426702,43426822,43427118,43427446,43427613,43428356,43428557,43438585,43441642],"created_at":"2025-03-20T17:27:46Z","created_at_i":1742491666,"num_comments":15,"objectID":"43426164","points":64,"story_id":43426164,"story_text":"Hi HN! I\u2019m Tony, co-founder of Inngest. I wanted to share AgentKit, our Typescript multi-agent library we\u2019ve been cooking and testing with some early users in prod for months.
Although OpenAI\u2019s Agents SDK has been launched since, we think an Agent framework should offer more deterministic and flexible routing, work with multiple model providers, embrace MCP (for rich tooling), and support the unstoppable and growing community of TypeScript AI developers by enabling a smooth transition to production use cases.
This is why we are building AgentKit, and we\u2019re really excited about it for a few reasons:
Firstly, it\u2019s simple. We embrace KISS principles brought by Anthropic and HuggingFace by allowing you to gradually add autonomy to your AgentKit program using primitives:
- Agents: LLM calls that can be combined with prompts, tools, and MCP native support.
- Networks: a simple way to get Agents to collaborate with a shared State, including handoff.
- State: combines conversation history with a fully typed state machine, used in routing.
- Routers: where the autonomy lives, from code-based to LLM-based (ex: ReAct) orchestration
The routers are where the magic happens, and allow you to build deterministic, reliable, testable agents.
AgentKit routing works as follows: the network calls itself in a loop, inspecting the State to determine which agents to call next using a router. The returned agent runs, then optionally updates state data using its tools. On the next loop, the network inspects state data and conversation history, and determines which new agent to run.
This fully typed state machine routing allows you to deterministically build agents using any of the effective agent patterns \u2014 which means your code is easy to read, edit, understand, and debug.
This also makes handoff incredibly easy: you define when agents should hand off to each other using regular code and state (or by calling an LLM in the router for AI-based routing). This is similar to the OpenAI Agents SDK but easier to manage, plan, and build.
Then comes the local development and moving to production capabilities.
AgentKit is compatible with Inngest\u2019s tooling, meaning that you can test agents using Inngest\u2019s local DevServer, which provides traces, inputs, outputs, replay, tool, and MCP inputs and outputs, and (soon) a step-over debugger so that you can easily understand and visually see what's happening in the agent loop.
In production, you can also optionally combine AgentKit with Inngest for fault-tolerant execution. Each agent\u2019s LLM call is wrapped in a step, and tools can use multiple steps to incorporate things like human-in-the-loop. This gives you native orchestration, observability, and out-of-the-box scale.
You will find the documentation as an example of an AgentKit SWE-bench and multiple Coding Agent examples.
It\u2019s fully open-source under the Apache 2 license.
If you want to get started:
- npm: npm i @inngest/agent-kit
- GitHub: https://github.com/inngest/agent-kit
- Docs: https://agentkit.inngest.com/overview
We\u2019re excited to finally launch AgentKit; let us know what you think!","title":"Show HN: AgentKit \u2013 JavaScript Alternative to OpenAI Agents SDK with Native MCP","updated_at":"2026-01-11T10:14:57Z","url":"https://github.com/inngest/agent-kit"}],"hitsPerPage":20,"nbHits":361,"nbPages":19,"page":0,"params":"query=Agentkind&advancedSyntax=true&analyticsTags=backend","processingTimeMS":14,"processingTimingsMS":{"_request":{"roundTrip":15},"afterFetch":{"format":{"total":1},"merge":{"mergeLoop":{"prepareNextHit":3,"total":3},"total":3},"total":3},"fetch":{"query":8,"scanning":1,"total":10},"total":14},"query":"Agentkind","serverTimeMS":16}