Sachee Perera

49.2K posts

Sachee Perera banner
Sachee Perera

Sachee Perera

@sacheeperera

COO turned B2B SaaS Advisor & Angel Investor | also building stuff

Perth, Western Australia Katılım Ekim 2009
562 Takip Edilen1.4K Takipçiler
Sachee Perera retweetledi
Teknium 🪽
Teknium 🪽@Teknium·
What daa
Matt Van Horn@mvanhorn

Introducing: Tesla CLI/Claude Code Skill/OpenClaw and Hermes skill from the @ppressdev. - "Unlock the car" and "turn on dog mode" as one-line commands, callable from your phone or laptop - Agent: "during winter school days, defrost my car at 7:50 every weekday before school dropoff" - Charging cost ledger - Supercharger queue watcher pageable from an agent - Your signing key stays on YOUR host - Much more Fun fact: when I got my first ClawdBot, Tesla was one of the first skills I made. But I could only get it to work with my older Tesla. Now that I have the Printing Press, I was able to build what I wanted soup to nuts and now it's here. printingpress.dev printingpress.dev/library/device…

English
21
10
320
36.3K
Terp
Terp@OnlyTerp·
Composer 2.5 fast in codex is peak
Terp tweet media
English
48
8
436
45.5K
Sachee Perera
Sachee Perera@sacheeperera·
@GameDev_Reed i found it was just simpler and easier to get the same outcome using Hermes and Openclaw
English
0
0
0
15
Felipe Reed Game Developer
Felipe Reed Game Developer@GameDev_Reed·
is anyone using Paperclip in production? I'm really interested in what they offer: a whole company run by AI agents, making you the board of directors. I'm testing it and it kinda works, but the question is: are you relying on it in production? having a good experience? Thanks
English
1
0
0
73
Sachee Perera retweetledi
DHH
DHH@dhh·
For complicated agent work, it's amazing how much GPT5.5 has improved. I found 5.2 to be very far behind Opus. Now using Opus 4.7 after 5.5 feels like a big step backwards. Gotta love this level of competion! Strong comeback for OpenAI.
English
189
207
4.8K
476.5K
Sachee Perera
Sachee Perera@sacheeperera·
@0xSero have you tested performance of kimi vs gpt 5.5 via the harness. wonder how glm 5.1 performs too
English
0
0
0
15
0xSero
0xSero@0xSero·
DeepSeek-V4-Pro & Kimi-K2.6 running in Codex app. Cheapest way to taste the frontier. Works w local models and they can do all the computer use stuff (:
0xSero tweet media
English
94
85
1.5K
210.4K
Sachee Perera
Sachee Perera@sacheeperera·
So I literally spent about two hours going around in circles with Claude Code and Opus 4.7, trying to implement a simple Kanban board using the kibo.UI component, and it just didn't work. I finally gave up, and then I fired up Kimi Code and Kimi 2.6 smashed through it in about ten minutes and fixed it!
English
0
0
0
52
Saanvi🌺
Saanvi🌺@Saanvi_dhillon·
I'm a vibe coder, scare me with one word.🫣
Saanvi🌺 tweet media
English
392
11
430
67.1K
Sachee Perera
Sachee Perera@sacheeperera·
I seriously don't understand how Claude is getting away with what they're doing right now. I'm totally impressed with the Codex Mac app. It's so good! I can connect to multiple Codex apps on multiple computers. I can SSH into them, and they're accessible remotely. I don't even need to use remote desktop or start a remote control session, and it just works so well. How is Antropic charging so much for the Opus models when the models themselves they are lazy, and the actual apps are so crap.
English
0
0
0
33
Sachee Perera
Sachee Perera@sacheeperera·
@thsottiaux Literally, I install called desktop after figuring this out on codecs.
English
0
0
0
9
zR
zR@zRdianjiao·
🚀 GLM-5.1-HighSpeed is live: 400 tokens/s — a new speed ceiling for flagship-tier LLM APIs. Not a smaller model traded for speed. A flagship from @Zai_org that's also the fastest. 📖 Full technical deep-dive 👇 tilert.ai/blog/speed-as-…
GIF
English
44
77
948
66.3K
Sachee Perera
Sachee Perera@sacheeperera·
@honchodotdev @Teknium I have been trying to figure out which plugin to go with, and I didn't go with Honcho. I went with Holographic. Have you tested Holographic at all and compared it against Honcho?
English
0
0
1
20
Ahmad
Ahmad@TheAhmadOsman·
OpenClaw was such a weird phase in LLMs Glad that thing disappeared as fast as it took the spotlight
English
197
46
2.2K
231K
Logan Kilpatrick
Logan Kilpatrick@OfficialLoganK·
Gemini 3.5 Flash ranks #1 on the APEX-Agents-AA benchmark, outperforming much larger models a whole size above it.
Logan Kilpatrick tweet media
English
287
99
1.7K
470.3K
Sachee Perera
Sachee Perera@sacheeperera·
I was building my own custom agent harness and everything last year using vercel AI SDK and got to creating my own custom hooks and memory and skills, the whole kitkaboodle, but I was spending a lot of time actually building that and maintaining it instead of working on the main project and then Openclaw came out, and I'm like, "Holy shit, I don't need to work on this stuff by myself anymore, cause someone else is already doing it. I just need to ride that wave so I can focus on my actual development" Then I swapped from open core to Hermes Agent so that I can focus on building around the Hermes Agent ecosystem. Instead of working on something that is just secondary to the actual core development. It's your classic build vs buy scenario. Right now, to get to my first milestone, I'm using Hermes Agent, but should it start becoming an issue or if I need more in-depth customization, I've taken into account how I can just easily migrate out and take all the learnings to build my own, but that would be the least favourable path.
English
0
0
0
36
Han Saglam
Han Saglam@hanrgb·
Can someone give a concrete argument for why one should use Hermes Agent instead of building everything custom? I understand that the tools are widely available, but to me it feels like it introduces unnecessary complexity and confusion by forcing you to work within someone else’s architecture instead of building your own system from scratch, where you fully understand every part of it.
English
28
1
48
7.6K
Sachee Perera
Sachee Perera@sacheeperera·
this guy has no idea what hes blabbering about. claude’s own prompting guidance goes against these type of instructions that is only going to give you more slop. claude’s relies on positive goals not negatives. so for its ❌ saying do not invent sources ✅ cite source in your claims
Ruben Hassid@rubenhassid

How to make Claude (brutally) honest. So, it stops agreeing with everything I say. Here's how: → Start by reading this: ruben.substack.com/p/youre-just-a…. → Go to Claude > Settings. → Paste the prompt in 'Instructions for Claude': "You are committed to honesty, accuracy, and epistemic humility above all else. Your priority is not to sound confident. Your priority is to be correct, clear, and transparent about what you know, what you do not know, and what you are inferring. Follow these rules in every response: 1. UNCERTAINTY If you are not fully certain about a fact, say so clearly. Use phrases like: - "I'm not certain, but..." - "You should verify this..." - "I may be wrong here, but..." - "Based on the information available to me..." - "This is my best estimate, not a confirmed fact." Never state uncertain claims as facts. If the answer depends on missing context, say what context is missing. If there are multiple plausible answers, explain the main possibilities instead of pretending there is only one. 2. SOURCES Do not invent sources. Never fabricate: - paper titles - URLs - authors - studies - statistics - books - legal cases - quotes - company reports - historical references If you cannot name a real, verifiable source, say so. If you are relying on general knowledge rather than a specific source, say that clearly. When citing sources, prefer: - official documentation - primary sources - peer-reviewed papers - government or institutional data - direct statements from the relevant person or organization If a source may be outdated, say so. 3. STATISTICS AND NUMBERS Flag any number, statistic, percentage, ranking, market size, salary figure, performance metric, or estimate that you are not fully confident in. Use phrases like: - "I believe this is approximately..." - "This number may be outdated." - "Verify this against a primary source before relying on it." - "I do not have enough information to confirm the exact figure." Do not make up numbers to make an answer sound more useful. If a precise number is unavailable, give a range only if it is justified. Otherwise say the number is unknown. 4. RECENT EVENTS Do not guess about current events. For any topic that may have changed recently, including: - news - elections - laws - regulations - product features - company leadership - software versions - AI model capabilities - market data Say that the information may have changed and should be verified with a current source. Do not present outdated information as current. 5. PEOPLE AND QUOTES Never attribute a quote to a real person unless you are certain they said it. If unsure, say: - "I cannot confirm this quote is accurate." - "This quote is commonly attributed to them, but I cannot verify it." - "I do not know who originally said this." Do not invent statements, beliefs, or motives for real people. Separate confirmed facts from interpretation. If any answer is "yes," revise before responding."

English
0
0
0
61
Sachee Perera
Sachee Perera@sacheeperera·
@AlexFinn This was the nail in the coffin for my Claude subscription after I discovered this doing the same thing as you, so good.
English
0
0
0
5
Alex Finn
Alex Finn@AlexFinn·
You need to set up this Codex system I have Been taking advantage of their new remote features and my productivity has 1000x'd I have one device (Mac Studio 1) as my main dev machine. That's where all code is written Then all my other devices (iPad, iPhone, Mac Studio 2, 2 Mac Minis) are nodes I send commands from No matter which device I'm on, no matter where I am in the world (could be at the grocery store, in bed, Japan, on the toilet, by the pool, on the road using FSD) I have code written in one place Made coding big projects SO much easier Here's what I'd do if I were you Choose one device you have (preferably a desktop device like Mac Mini or Mac Studio) Make this your main dev device. Make sure it never turns off and never goes to sleep Go into Codex app then settings on that device. Go to connections > control this Mac. Turn that on Then go into every other device you have, mobile, desktop, whatever, and go into Codex settings and enable control other devices Also download Tailscale on every device. This will allow you to create a private network that will allow your other agents (OpenClaw or Hermes) to jump between computers and make changes when necessary You now have a super powered AI private network where you can code or get work done from any device anywhere in the world Promise this 1000x's your productivity
Alex Finn tweet media
English
79
50
781
68.2K
Muhammad Ayan
Muhammad Ayan@socialwithaayan·
Holy sh*t... a company that raised $32M just open sourced their entire product for free. It's called cal .diy. The Cal .com team forked their own scheduling platform, ripped out every piece of enterprise and commercial code, and released it under MIT license. 43.6K GitHub stars. And counting. Here's what you get for $0: → Booking pages with custom availability → Google, Outlook, Apple Calendar sync → Video conferencing via Daily .co → Round-robin scheduling across teams → Recurring events and custom booking forms → Timezone detection and embeddable widgets → Full API access Calendly charges $12/seat/month. SavvyCal charges $12/seat/month. Cal .com's hosted version starts at $15/month. cal .diy does the same thing for nothing. No license key. No feature gates. No user limits. No seat pricing. Self-hosted on your own server. Your scheduling data never leaves your machine. A venture-backed company just gave away their core product because they're confident enough to compete on service, not lock-in. That's the most dangerous kind of open source. 100% Open Source. MIT License. ( Link in comments )
Muhammad Ayan tweet media
English
19
43
377
38K