Arun Tripathi

Forever student 📚 |Thinker 👽| Exploring the past, understanding the present, shaping the future 🌍📜 | Tech enthusiast 🖥️Nationalist

[COPWATCH INDIA @copwatchIN]

Latur, Maharashtra Police !

Are we waiting for people to die at the hands of such Police Officers ?

When are we going to say enough is enough ?

The role of the traffic police is to regulate traffic, enforce road safety, and issue penalties (challans) for violations.

They do not have the authority to inflict physical punishment on citizens. Any act of physical violence by an officer is a violation of the law and human rights.

Right to Due Process: If a rider fails to stop, the legal recourse for the police is to note the vehicle's registration number, track the owner, and issue a challan or summon the individual to court.

[Ghar Ke Kalesh @gharkekalesh]

Police stopped a man returning from a movie with wife & kids, took his bike keys & slapped him on the road despite showing ticket.

[Ronin @DeRonin_]

Andrej Karpathy: "90% of Claude's mistakes come from missing context, not a weak model."

41% mistake rate without a CLAUDE.md. 11% with the 4-rule baseline. 3% with the 12-rule version below

here are the 12 rules senior engineers settled on:

1. think before coding: state assumptions, don't guess. the model can't read your mind, stop hoping it will

2. simplicity first: minimum code, no speculative abstractions. the moment you let Claude add "for future flexibility," you've added 200 lines you'll delete next quarter

3. surgical changes: touch only what you must. don't let it improve adjacent code, that's how PRs blow up

4. goal-driven execution: define success criteria upfront, loop until verified. without them Claude either loops forever or stops too early

5. use the model only for judgment calls: classification, drafting, summarization, extraction. NOT routing, retries, status-code handling, deterministic transforms. if code can answer, code answers

6. token budgets are not advisory: per-task 4000, per-session 30000. by message 40 of a long debug, Claude is re-suggesting fixes you rejected at message 5

7. surface conflicts, don't average them: two patterns in the codebase? pick one. Claude blending them is how errors get swallowed twice

8. read before you write: read exports, callers, shared utilities. Claude will happily add a duplicate function next to an identical one it never read

9. tests verify intent, not just behavior: a test that can't fail when business logic changes is wrong. all 12 of Claude's tests can pass while the function returns a constant

10. checkpoint every significant step: Claude finished steps 5 and 6 on top of a broken state from step 4. nobody noticed for an hour

11. match the codebase conventions: class components? don't fork to hooks silently. testing patterns assumed componentDidMount, hooks broke them without surfacing

12. fail loud: "completed successfully" with 14% of records silently skipped is the worst class of bug. surface uncertainty, don't hide it

what actually compounds instead of the next framework:

- the CLAUDE.md file as institutional memory across sessions
- eval-driven changes, not vibe-driven
- checkpoints over speed
- explicit conflicts over silent blending
- discipline over framework, every time
- one repo, one rules file, no exceptions

be a few rules ahead of AI twitter before this becomes mass-opinion

study this

[Yogi Adityanath @myogiadityanath]

नमाज पढ़नी है, आप शिफ्ट में पढ़िए...

प्यार से मानेंगे ठीक है, नहीं मानेंगे तो दूसरा तरीका अपनाएंगे...

[Mohandas Pai @TVMohandasPai]

Folks this is a great temple, please visit. @LalithHotel there is a great place to stay

[Ajeet Bharti @ajeetbharti]

प्रिय @BJP4India @dpradhanbjp @narendramodi, NEET समेत हर परीक्षा के पेपर लीक के कारण मरते बच्चे के दोषी आप सब हो। गोवा के बाद आज उत्तर प्रदेश में रितिक मिश्रा ने तनाव में आ कर आत्महत्या कर ली।

ईश्वर न करे एक भी जीवन ऐसे नकारे मंत्रियों के कारण जाए, पर वास्तविकता यह है कि मैं जानता हूँ यह संख्या बढ़ेगी ही।

[Sanjay Dixit ಸಂಜಯ್ ದೀಕ್ಷಿತ್ संजय दीक्षित @Sanjay_Dixit]

He has said this outside the House, so he can be and should be prosecuted. Request @MRVChennai to do the needful.

[The Hawk Eye @thehawkeyex]

The point is not what he said.
The point is - there were ~200 Hindu MLAs in the assembly. None objected.

[Arun Pudur @arunpudur]

⚡ Joseph Vijay appoints a Muslim as Minister for Hindu Religious and Charitable Endowments (HRCE).

Utter destruction and usurping of Hindu temples 2.0 has begun.

As predicted TVK is DMK 2.0.

[The Jaipur Dialogues @JaipurDialogues]

Hearing News of Cabinet Reshuffle of Modi Government

Bhai sabse pehle Dharmendra Pradhan ko bahar karo!

[Ajeet Bharti @ajeetbharti]

Fuck you Dharmendra Pradhan!

[Zack Korman @ZackKorman]

Mythos found a single vulnerability in cURL (along with three false positives, and one issue they classified as a bug). The founder/lead dev wasn't super impressed.

[Avid @Av1dlive]

Garry Tan (CEO of Y-Combinator): "when someone asks how I 'prompt' my AI, the answer is: I don't. the skills are the prompts."

[if I had 7 days to master skills and how to use them to automate workflows:]

→ read the Skillify 11-item checklist (SKILL.md in gbrain)
→ watch Murag + Barry Zhang: "Don't Build Agents. Build Skills Instead."
→ Read "Designing, Refining, and Maintaining Agent Skills at Perplexity"
→ do one workflow. type /skillify. watch it become permanent.

that's the whole day.

here is how to set it up:

1. clone GBrain (his open-source second brain, Postgres-backed memory + 30 skills)

2. add GStack (23 battle-tested slash-command skills, drops right in)

3. do anything once → type /skillify → it's a skill forever

prompting is dead. skillifying is next.

[BJP @BJP4India]

Udhayanidhi Stalin has once again spread venom against Sanatan Dharma. He has again called for Sanatan to be eradicated.

Remember, he is the same person who compared Sanatan Dharma to diseases such as malaria and dengue.

𝐖𝐞𝐥𝐥, 𝐭𝐡𝐞 𝐩𝐞𝐨𝐩𝐥𝐞 𝐨𝐟 𝐓𝐚𝐦𝐢𝐥 𝐍𝐚𝐝𝐮 𝐡𝐚𝐯𝐞 𝐚𝐥𝐫𝐞𝐚𝐝𝐲 𝐫𝐞𝐣𝐞𝐜𝐭𝐞𝐝 𝐡𝐢𝐦 𝐚𝐧𝐝 𝐡𝐢𝐬 𝐢𝐝𝐞𝐨𝐥𝐨𝐠𝐲… 𝐒𝐨 𝐲𝐨𝐮 𝐤𝐧𝐨𝐰 𝐭𝐡𝐞 𝐦𝐨𝐨𝐝 𝐨𝐟 𝐭𝐡𝐞 𝐬𝐭𝐚𝐭𝐞.

[Gems @gemsofbabus_]

🚨 Madhya Pradesh Textbook Corporation’s new Chairman arrived in Bhopal from Ujjain with a massive convoy (>500 vehicles) to assume charge.

Meanwhile, the Prime Minister tells citizens and the CM tells ministers: “Save Petrol.” 🤡🤦‍♂️

[Priyanka Chaturvedi🇮🇳 @priyankac19]

2024- NEET UG paper leak, protests, matter reached Supreme Court
2025 - UGC Guidelines, protests, matter in Supreme Court
2026 - NEET UG paper leak again, students upset, Exam rescheduled.

NTA: Irresponsible
Education Minister : Unchanged
Accountability: Zero

[Anatoli Kopadze @AnatoliKopadze]

This 3-minute video from Anthropic will make you realize you've been using Claude wrong this whole time

"when you open Claude Code without a CLAUDE.md file, it has to start fresh every single time"

most people have no idea how much this one file changes the way Claude actually works for you

and this isn't only for developers

the article below has 21 CLAUDE.md examples you can use in your daily routine

I'm sure everyone will find something that fits exactly how they work

[Priyanka Chaturvedi🇮🇳 @priyankac19]

हमारा धर्म अमर है क्योंकि यह किसी पुस्तक या पैगंबर पर नहीं, बल्कि आत्मा के उन शाश्वत नियमों पर आधारित है जो कल भी सत्य थे और आज भी हैं ~ स्वामी विवेकानंद

[Akash Dixit | ಆಕಾಶ್ ದೀಕ್ಷಿತ್ @dixithru]

Stop saying that i am from South India or i am from North India, Instead say that you are from the South Part of India or the North part of India. This small correction will make a difference and it unifies the Nation.

- Jagadguru Sri Sri Vidhushekhara Bharati Sannidhana.

[Akshay 🚀 @akshay_pachaar]

As an AI Engineer. Please learn:

- Harness engineering, not just prompt engineering
- Prompt caching vs. semantic caching tradeoffs
- KV cache management at scale
- Speculative decoding vs quantization
- Structured output failures & fallback chains
- Evals (LLM-as-judge + human evals)
- Cost attribution per feature, not just per model
- Agent guardrails & loop budgets
- LLM observability as a first-class discipline
- Model routing & graceful fallback logic
- Knowing when to fine-tune vs. in-context learning