Paul-André Deleu @padeleu

I teamed up with @Liz_Bayley_Physio to create this progressive rehab programme for Plantar Heel Pain.  We've called it 'LiTHE' (Liz and Tom's Heel Exercises) and it's aimed at active and sporty patients. It's not a recipe though so it's best to adapt it to suit individual needs.

We removed ~80% of the Claude Code system prompt for our newest models, this is what we've learned about writing system prompts, skills and Claude.MDs for them. x.com/i/article/2080…

x.com/i/article/2057…

Anthropic pays $750,000+ a year for engineers who can build LLM architectures from scratch. Stanford taught the entire thing in 1 hour lecture & released it for free. Bookmark & watch this today before someone takes it down ...

x.com/i/article/2066…

Andrew Ng just dropped the best 2-hour prompt engineering course of 2026 From brainstorming and agentic prompt patterns To building prompts for images, video, apps, and data analysis: 00:00 - introduction to prompt engineering 38:12 - brainstorming prompt patterns 47:59 - context engineering 01:30:27 - prompts for media generation 01:53:43 - app building and data analysis prompts This 2-hour watch will change how you use LLMs Bookmark and watch it today Then read the article below

rari @0xwhrrari

3 months ago

x.com/i/article/2065…

57 194 1K 3.6M 4K

🚨 Anthropic just showed a 27-minute workshop on how to actually do prompts for Claude. Taught by the people who built it. Free. No registration. No paywall. I've seen $300 courses that don't cover what they teach in the first 8 minutes. Watch it and bookmark it now.

A PhD student at Oxford got caught submitting "AI-generated" work. Except he hadn't used AI to write anything. He used it to think. Here's the workflow his advisor called "the most sophisticated research process I've seen in 20 years." He starts every essay with a brutal diagnostic prompt. Dumps his rough argument into Claude and asks: "What are the 3 weakest logical jumps in this reasoning? Where would a hostile examiner attack first?" The AI doesn't write his essay. It destroys his draft. Then he rebuilds. But the next step is what separates him from every other student using ChatGPT or Claude to generate paragraphs. He uploads the top 5 papers in his field and asks: "What claims in my argument contradict or oversimplify what these authors actually found?" Most students cite papers they've skimmed. He cites papers he's been forced to genuinely understand. The final move is almost unfair. Before submitting, he pastes his conclusion and asks: "What would a philosopher of science say is missing from this argument? What assumptions am I making that I haven't defended?" His essays come back with comments like "unusually rigorous" and "demonstrates rare critical depth." He's not using AI to write. He's using it to think harder than he could alone. The tool hasn't changed. The workflow has.

🦶 strengthening very nice study: leaning forward and adding mass ⬆️⬆️foot muscles activity and torque output Add this for richer foot-ankle strength work🔥🏋️‍♀️ Totally in line with our practice and recent training study proposal 📄sciencedirect.com/science/articl…

keyboard_arrow_left Previous keyboard_arrow_right Next

journals.lww.com/painrpts/fullt… Pretty important stuff! "Social determinants of health equity for people with pain: from understanding to action"

jamanetwork.com/journals/jaman… So pain is “in the brain”🧠 😳😬 Do you think this can be confidently said? ”maladaptive plasticity in the brain can generate persistent pain independent of any peripheral nociceptive input”

Payers, insurers, and employers: take a proactive approach to prevent diabetes and reduce health care costs. Read more on the DPA’s Six Pillars of Value. professional.diabetes.org/clinical-suppo…

Excellent systematic review of biomechanics & running economy from @BasVanHooren & colleagues: link.springer.com/article/10.100… Increasing step rate or reducing trunk flexion may help efficiency in some runners 🏃‍♂️ Lots to unpack so we’ve summarised the key findings in this graphic.

FHL & ankle DF 🦶🏼 Mini thread 🧵 🔹 FHL tightness was found in almost half of 188 Pts - it’s more common than previously recognised 🔹Ankle DF ⬇️ as FHL tension ⬆️ in ‘normal’ and symptomatic Pts 🔹A tight FHL can be linked to hallux rigidus/dorsal osteophytes 1/5

2 exercises to target🦶🏼strength without MTPJ flex/ext. Great for: 1st toe OA/Gout Turf toe Plantar plate tear Capsulitis/synovitis Freiburgs Sesamoiditis There’re lots of other important Rx options for metatarsalgia, but avoiding secondary weakness can help improve outcomes 🏋🏼‍♀️

In case you’re wondering what a subluxing peroneal tendon looks/sounds like… 😯🦶🏼👀 #SoundOn 🔈 This is the result of multiple sprains in a gymnast, who was required to continue training/performing with an acute injury… She’s also extremely hypermobile #ItsNeverJustASprain

An Ankle Sprain is also a Brain Sprain 🙏

keyboard_arrow_left Previous keyboard_arrow_right Next

FHL stress test: Measures FHL excursion: Ankle in maximal dorsiflexion, stabilise the 1st metatarsal and maximally dorsiflex the hallux MTP joint: Normal >20° dorsiflexion Tight <20° dorsiflexion 2/5

Take aways 🧵1/3 Functional foot exercises: 🦶🏼 🔹SLHR 🔹Toe walking 🔹Hopping These ex’s take advantage of the PF’s neural drive, include balance + W/B, and take toes into extension - all of which bias them to ⬆️ activation. Plus adherence can be better… pubmed.ncbi.nlm.nih.gov/37126537/

Positive knee to wall FHL test in this gymnast with posteromedial ankle pain. Difference of 4cms with the big toe extended. No change on uninjured side. Michelson et al (2021) say this test isn’t sensitive enough to pick up a tight FHL… Continued 1/2 🦶🏼 pubmed.ncbi.nlm.nih.gov/32739176/