•twitter15877
listening to
timeline
15877
mentions
tracked
tracked
net sentiment
+24%
4998 positive
9734 neutral
1145 negative
volume & sentiment over time
Sep 30 – Oct 14, 2026 · Daily
sources
most discussed
—
active days
15
mentions
Showing 2876–2900 of 4998 mentions
-
• twitter 1w positiveBookmark this Hermes command cheat sheet for your next session.🔖 The new edition has a fresh layout, updated command guidance, and badges showing where commands work across Desktop, CLI, messaging, and the shell. It covers useful controls like: - /goal - /loop - /heartbeat - /bg - /btw - /plan - /review - /refine - … Bookmark this Hermes command cheat sheet for your next session.🔖 The new edition has a fresh layout, updated command guidance, and badges showing where commands work across Desktop, CLI, messaging, and the shell. It covers useful controls like: - /goal - /loop - /heartbeat - /bg - /btw - /plan - /review - /refine - /moa - /busy This edition also adds hermes pause and hermes resume to Terminal Essentials, expands the CLI queue guidance, and clarifies branching and saved sessions. The sheet is organized around what you’re trying to do: - Sessions & context - Active work - Goals, loops & plans - Models & behavior - Skills, memory & tools - Automation & coordination - Inspection, recovery & troubleshooting Messaging/Gateway commands and Terminal Essentials are included, too. Command names, syntax, and availability checked against Hermes source. Keep it nearby when you need to queue the next task, steer a running job, or figure out what went wrong.view source ↗
-
• twitter 1w positivegoogle just released nano banana 2.1 model is looking good but don't think it's as good as gpt images it also failed at creating an ending for avengers doomsday, but gpt images was able to make it here's one comparison of a kpop group, just look at the difference https://t.co/bmgZhsQHye google just released nano banana 2.1 model is looking good but don't think it's as good as gpt images it also failed at creating an ending for avengers doomsday, but gpt images was able to make it here's one comparison of a kpop group, just look at the difference https://t.co/bmgZhsQHyeview source ↗
-
• twitter 1w positive"I can't talk to girls" lmao this is a great one https://t.co/lrGobSzZie "I can't talk to girls" lmao this is a great one https://t.co/lrGobSzZieview source ↗
-
• twitter 1w positiveAnother Devin Max Plan for Free, don't miss out! 10 Max plans to grab. https://t.co/Ed3qrI6rEK Another Devin Max Plan for Free, don't miss out! 10 Max plans to grab. https://t.co/Ed3qrI6rEKview source ↗
-
• twitter 1w positiveTern is the best agent multiplexer ever. It's made by the creator of OMP, extremely fast and looks amazing. Tern is in closed beta but it's already MILES ahead its competition and keeps improving very quickly. https://t.co/roI07coaj8 Tern is the best agent multiplexer ever. It's made by the creator of OMP, extremely fast and looks amazing. Tern is in closed beta but it's already MILES ahead its competition and keeps improving very quickly. https://t.co/roI07coaj8view source ↗
-
• twitter 1w positiveMy givaway is over guys, don't submit anymore this one is active! Try to grab a month of Devin Max for free! My givaway is over guys, don't submit anymore this one is active! Try to grab a month of Devin Max for free!view source ↗
-
• twitter 1w positiveguess the less I code, the better I get at TikTok https://t.co/8a173VbnxD guess the less I code, the better I get at TikTok https://t.co/8a173VbnxDview source ↗
-
• twitter 1w positive@usgraphics I feel like this happened pre-AI tbh @usgraphics I feel like this happened pre-AI tbhview source ↗
-
• twitter 1w positiveI like that we go back to idea for cost-efficient specialized models I am tired to hear 'but my claude can already to that', sure it can do everything. But you don't know for what price because you don't pay for it. Your employer pays! And token cost if what matter to business I like that we go back to idea for cost-efficient specialized models I am tired to hear 'but my claude can already to that', sure it can do everything. But you don't know for what price because you don't pay for it. Your employer pays! And token cost if what matter to businessview source ↗
-
• twitter 1w positiveI asked Opus 5.5 to fix one of my biggest annoyances on Github (clicking screenshots opens in new tab) 1.5 minutes later, I had a working Chrome extension that fixes the behavior. I know this is basic stuff, but it still feels magical that you can just fix things. https://t.co/mkslLT9Vp3 I asked Opus 5.5 to fix one of my biggest annoyances on Github (clicking screenshots opens in new tab) 1.5 minutes later, I had a working Chrome extension that fixes the behavior. I know this is basic stuff, but it still feels magical that you can just fix things. https://t.co/mkslLT9Vp3view source ↗
-
• twitter 1w positiveLove getting these messages https://t.co/4xQZTxzWOa Love getting these messages https://t.co/4xQZTxzWOaview source ↗
-
• twitter 1w positiveI'm certain that Instinct will defeat Muse. The instinct product philosophy feels very apple like, crisp, clear & benefits driven. Muse is so distracted and almost Samsung/Android energy. It feels like Muse is merely building a better tool. Instinct is thinking network effects and owning the 'personal context graph'. … I'm certain that Instinct will defeat Muse. The instinct product philosophy feels very apple like, crisp, clear & benefits driven. Muse is so distracted and almost Samsung/Android energy. It feels like Muse is merely building a better tool. Instinct is thinking network effects and owning the 'personal context graph'. very bullishview source ↗
-
• twitter 1w positiveSpaceXAI will run four free live Grok Bot workshops on Zoom this week, October 6-9. Martin Lacsamana opens with Grok Bot 101 on Tuesday. Jenna Nanpei covers marketing workflows on Wednesday. The people building Grok Bot walk through what’s new on Thursday. Joseph Yang closes with Grok Bot 201 on Friday. Each session … SpaceXAI will run four free live Grok Bot workshops on Zoom this week, October 6-9. Martin Lacsamana opens with Grok Bot 101 on Tuesday. Jenna Nanpei covers marketing workflows on Wednesday. The people building Grok Bot walk through what’s new on Thursday. Joseph Yang closes with Grok Bot 201 on Friday. Each session runs 10-11 a.m. PT. Registration is free, and registrants who miss the live session get the recording. 👇view source ↗
-
• twitter 1w positivethank you everyone for 1,000+ followers! i'll be sharing my guide on how to speedrun growth on X tomorrow morning :) as promised, i'll get the keyboard giveaway going soon if you're seeing this & we're not following each other, let's connect! 👋 https://t.co/JpCCxoQhJg thank you everyone for 1,000+ followers! i'll be sharing my guide on how to speedrun growth on X tomorrow morning :) as promised, i'll get the keyboard giveaway going soon if you're seeing this & we're not following each other, let's connect! 👋 https://t.co/JpCCxoQhJgview source ↗
-
• twitter 1w positiveTomorrow will be a better day. 🌞🌱 https://t.co/1pR7YACNPg Tomorrow will be a better day. 🌞🌱 https://t.co/1pR7YACNPgview source ↗
-
• twitter 1w positivebrah i just realized i look like the nous research girl https://t.co/r40MO32sPV brah i just realized i look like the nous research girl https://t.co/r40MO32sPVview source ↗
-
• twitter 1w positiveI redesigned the vibe-coded product UI, and it turns out to be something genuinely promising. https://t.co/WhMQE4BG4R I redesigned the vibe-coded product UI, and it turns out to be something genuinely promising. https://t.co/WhMQE4BG4Rview source ↗
-
• twitter 1w positiveAnthropic used Claude to make their apps 3x faster. The strategies they came up with to do it are fascinating. I broke it down so you can use the best parts in your own work :) https://t.co/6utQFWcLJp https://t.co/lQRwvAtLEe Anthropic used Claude to make their apps 3x faster. The strategies they came up with to do it are fascinating. I broke it down so you can use the best parts in your own work :) https://t.co/6utQFWcLJp https://t.co/lQRwvAtLEeview source ↗
-
• twitter 1w positive50% speed increase....for free?? 50% speed increase....for free??view source ↗
-
• twitter 1w positiveSpace Bunny Alpha vs GPT-6.1 Sol Ultra, The quality from this stealth model is seriously impressive, and at this point almost everything seems to point to it being MiniMax 3.1, It took around 5x longer than GPT-6.1 Sol, but mostly because it kept retrying and refining the work until it actually satisfied the full promp… Space Bunny Alpha vs GPT-6.1 Sol Ultra, The quality from this stealth model is seriously impressive, and at this point almost everything seems to point to it being MiniMax 3.1, It took around 5x longer than GPT-6.1 Sol, but mostly because it kept retrying and refining the work until it actually satisfied the full prompt, The final result was clearly better I also ran the exact same prompt with Opus 5.5, and so far I’d honestly put Space Bunny Alpha first. Maybe we’ll finally find out which lab is behind it this week, but whoever built it has optimized it incredibly well for following instructions all the way through instead of stopping halfway, It’s been a long time since I’ve been this excited about a free stealth modelview source ↗
-
• twitter 1w positiveSee what a better edit can do? Want the same for your content? DM me. 📩🚀 https://t.co/btTGSBQl9u See what a better edit can do? Want the same for your content? DM me. 📩🚀 https://t.co/btTGSBQl9uview source ↗
-
• twitter 1w positiveOk so has anyone I mean anyone gotten anything usefull out of their dot? Please show me what its good for Ok so has anyone I mean anyone gotten anything usefull out of their dot? Please show me what its good forview source ↗
-
• twitter 1w positiveAleph Alpha’s Kolibri-1 is impressive. 🇩🇪 >Rent an H200 with 141 GB VRAM for ~$3.59/hour. >$20 gets you roughly 5.5 hours enough for a serious repo-level test. Spin up Kolibri-1 with vLLM, connect it to OpenHands as the agent harness, and throw a real codebase at it. 78B parameters! Only 3.46B active per token 262k … Aleph Alpha’s Kolibri-1 is impressive. 🇩🇪 >Rent an H200 with 141 GB VRAM for ~$3.59/hour. >$20 gets you roughly 5.5 hours enough for a serious repo-level test. Spin up Kolibri-1 with vLLM, connect it to OpenHands as the agent harness, and throw a real codebase at it. 78B parameters! Only 3.46B active per token 262k context Reasoning Tool calling Apache 2.0 Why should anyone care? Because you can run the weights yourself. If you want to keep sensitive code and data inside your own infrastructure or customize the stack or maybe control deployment and avoid being completely dependent on a closed-model API It’s also interesting because it’s also genuinely built for German + English, with a German-optimized tokenizer and substantial native German training data not simply an English-first model with German bolted on afterward I started testing it tonight and first impression: the speed-to-capability ratio looks very promising Now testing what I actually care about: Can an open-weight European model hold its own in a real agentic coding workflow?view source ↗
-
• twitter 1w positiveBackground subagents in the Cursor SDK now report back to the parent. Their results return as a follow-up turn on the same run, so stream() and wait() carry through until every subagent finishes. Custom tools can also carry MCP annotations like readOnlyHint and destructiveHint, so the model can tell a lookup from a de… Background subagents in the Cursor SDK now report back to the parent. Their results return as a follow-up turn on the same run, so stream() and wait() carry through until every subagent finishes. Custom tools can also carry MCP annotations like readOnlyHint and destructiveHint, so the model can tell a lookup from a delete before it calls one. You can replace Cursor’s system prompt with your own, and rules, skills and tool schemas still load. Cursor is enabling that account by account.view source ↗
-
• twitter 1w positive“Value” Meh. No. OpenAi can just pump api prices and win the benchmark. Objectively, you get a shitton of work done with line 6 models and Anthropic has no answer to Luna. Perhaps Anthropic api prices are just overinflated. I still appreciate a touch of genius from Opus models from time to time, but OpenAi prov… “Value” Meh. No. OpenAi can just pump api prices and win the benchmark. Objectively, you get a shitton of work done with line 6 models and Anthropic has no answer to Luna. Perhaps Anthropic api prices are just overinflated. I still appreciate a touch of genius from Opus models from time to time, but OpenAi provides true worker modelsview source ↗