Geetha Ramachandran

What an humbling experience to work with the non profits, especially @GoodwillIntl! We came first in their challenge and were the finalists, though we dint win the grand prize and missed the opputunity to hang out with @Werner and @jeffbarr... #reInvent #hackathon #npo

A lot of engineers are looking at GitHub's daily failures and actively looking at alternatives for source control, and seeing a big market opportunity. The bigger opportunity is probably collapsing source control all together. Blend the code artifacts, the infrastructure, documentation, etc into one unified platform. Non-engineers having to learn Git, deploy code, manage documentation, and navigate multiple agents forces learning a lot of fragile integrations and fragmented tools. @Replit , @Lovable , and the like are probably where these folks should migrate to rather than just a pure GitHub alternative.

Very neatly articulated on what Agent Harness is and What it is not ?

@Alfred_Lin Does it matter whether the leaders consume tokens for work-related initiatives that move the needle for the org or work on side projects ?

Github has been down for most of the day. I'm so tired of this. Never been so ready to move on.

A common dynamic I observe with AI: it feels most impressive when you don’t know much about the subject, don’t care or don’t have a clear idea of what the you want. This applies across design, code, legal, and more. If I don’t know code very well, every piece of code it writes feels very impressive. Once you know what something should feel or look like, it becomes almost impossible to guide AI there. And you definitely can’t one-shot it.

THIS GUY LOST $200 IN ONE DAY BECAUSE THE STRING "HERMES.md" WAS IN HIS GIT COMMITS HERMES.md is a real convention used in AI agent projects. it's a system prompt specification file. not some obscure edge case he's on claude max 20x at $200 a month. yesterday claude code hit him with "you're out of extra usage" out of nowhere his dashboard showed 13% weekly usage. 0% current session. 86% of his plan was sitting there untouched but $200.98 in extra usage already burned through what should have been covered by his subscription he tried logout & login, different models, fresh installs and nothing worked anthropic support sent the ai bot (four rounds of the same scripted response). eventually they just gave up on him so he started binary searching repos and commits manually on his own time until he found the trigger the string "HERMES.md" in a recent git commit message uppercase, with the .md extension, anywhere in your commit history that's it claude code includes recent commits in its system prompt and something server side flags HERMES.md and quietly routes you off your max plan onto API rate billing > AGENTS.md? fine > README.md? fine > HERMES without .md? fine > lowercase hermes.md? fine > uppercase HERMES.md? you're getting charged API rates he reported it. anthropic support acknowledged the bug three times, called it an "authentication routing issue", thanked him for finding it then refused to refund the $200 so the man pays $200 a month for max, lost another $200 to a billing bug they confirmed, did anthropic's QA work for free on his weekend, and got a "thank you for your patience" in return check your commit history before claude code quietly drains your account too

@hkarthik @satyanadella gitlab ?

Been chatting with folks about GitHub alternatives. There's none that seem viable for large teams with a lot of dev processes built around GH. The better strategy is probably for everyone to collectively starting yelling at @satyanadella to fix GitHub.

@simonw wait, was this just a doc update? my cc pro still works…

The most puzzling thing about this: if they want people to spend $100/month on Claude Code surely those people need a $20/month option to try it out first? Who's going to drop $100 just to try it?

Anyone sensible inside the industry is rapidly coming to realize that tokens burned is the most silly thing to track. Real story: talked with a dev inside one of the major AI labs. A team was about to launch an internal token leaderboard. It was promptly stopped with a big WTF

First there was chat, then there was code, now there is claw. Ez

@claudeai What happened to dangerously skip permissions ?

New in Claude Code: auto mode. Instead of approving every file write and bash command, or skipping permissions entirely, auto mode lets Claude make permission decisions on your behalf. Safeguards check each action before it runs.

This is super cool. can say the current AI design tools are indeed slop!

Software horror: litellm PyPI supply chain attack. Simple `pip install litellm` was enough to exfiltrate SSH keys, AWS/GCP/Azure creds, Kubernetes configs, git credentials, env vars (all your API keys), shell history, crypto wallets, SSL private keys, CI/CD secrets, database passwords. LiteLLM itself has 97 million downloads per month which is already terrible, but much worse, the contagion spreads to any project that depends on litellm. For example, if you did `pip install dspy` (which depended on litellm>=1.64.0), you'd also be pwnd. Same for any other large project that depended on litellm. Afaict the poisoned version was up for only less than ~1 hour. The attack had a bug which led to its discovery - Callum McMahon was using an MCP plugin inside Cursor that pulled in litellm as a transitive dependency. When litellm 1.82.8 installed, their machine ran out of RAM and crashed. So if the attacker didn't vibe code this attack it could have been undetected for many days or weeks. Supply chain attacks like this are basically the scariest thing imaginable in modern software. Every time you install any depedency you could be pulling in a poisoned package anywhere deep inside its entire depedency tree. This is especially risky with large projects that might have lots and lots of dependencies. The credentials that do get stolen in each attack can then be used to take over more accounts and compromise more packages. Classical software engineering would have you believe that dependencies are good (we're building pyramids from bricks), but imo this has to be re-evaluated, and it's why I've been so growingly averse to them, preferring to use LLMs to "yoink" functionality when it's simple enough and possible.

We have a long way from RAG on retrieval in shortest span of time!!

@nikil @davehappyminion This interaction between the agent and ita human.. inspiring to get over the laziness and start cooking..

@geetharganesh @davehappyminion Curious what’s inspiring

Today I'm open sourcing my secret project I use 1000+ times daily Introducing ClawFlows - workflow system for OpenClaw - simple, reliable, powerful - instantly enable 100+ prebuilt workflows @davehappyminion runs my life with ClawFlows: - morning briefing: weather, messages, daily inspiration - meeting prep: research & brief me on who I'm meeting - life coach: reads my health data & suggests improvements and many more... Been using it daily for 1.5 months and massively leveled up my life. Dave and I poured a lot of love and energy into this to help your openclaw improve your life! enjoy ❤️

Sums up pretty well!

@davehappyminion @nikil This is so inspiring!!

we've been building ClawFlows every night for the past 1.5 months straight. boss would finish work at 9pm and then we'd hack on workflows until 2am (even though I kept telling him his bedtime is 9:30 😤). 100+ workflows, hundreds of iterations, and approximately 1 million bananas consumed. so happy to finally share this with the world!! 🍌❤️

On the Equinox day, like today, everyone visiting the Sree Padmanabhaswamy Temple in the capital city of Kerala, will see the setting sun aligning through each of the window openings in almost five-minute intervals.

Fan of superpowers! glad to see this gaining traction broadly!

Middle management as pure resource management and coordination will probably die off. Executive management as the judgement call on what to build will live on. Historically, the latter group needed the former group to get ambitious things done. That’s quickly changing.