Miguel Miranda retweeted
La que se nos viene encima.
Para lo bueno y para lo malo.
Cuánto de cada, yo no tengo ni idea.
Miguel Miranda retweeted
I highly recommend running this skill on your projects
vulnerability-scanner reviews your code for security vulnerabilities, including AI agents and MCP servers
What it does:
- Maps the attack surface and trust boundaries.
- Detects prompt injection and insecure agent tools
- Checks for exposed secrets and vulnerable dependencies
- Reviews MCP servers, skills, hooks, and CLAUDE.md files
- Identifies OWASP vulnerabilities and CI/CD security risks
- Verifies findings to reduce false positives
- Reports each issue with Attack, Impact, and Fix
Some results from testing:
- Automatically activated in 7/7 runs
- 0 false positives across 7 decoys
- Compatible with 6 coding agents
IMPORTANT: It’s read-only and never modifies your code
Install:
npx claude-code-templates@latest --skill security/vulnerability-scanner
aitmpl.com/component/skills/…
It carries the core design but still needs some work to be more flexible.
github.com/xikhar/papercraft
We shrank 1 GB of Wikipedia to 105.39 MB, losslessly.
Introducing tufazip: the best entry on the Large Text Compression Benchmark using a neural network as its sole predictor. Everything above it mixes a transformer with hand-built models.
Full post 👇
tufalabs.ai/research/tufazip
China open-sourced a model that reconstructs any scene in 3D from a regular video, in real-time.
one camera. no LiDAR. 10,000+ frames without breaking down.
just walk around with your camera and see the whole world rebuilt in 3D at 20 fps.
→ ~20 FPS on a single GPU
→ Stays stable across 10,000+ frames
→ Outperforms optimization-based methods on benchmarks
→ Handles driving footage, outdoor scenes and indoor walkthroughs
100% open source.
I copy any animation I see now, these 18 GitHub repos are the reason
a f*cking gold for Fable 5.2
Copy. Style. Paint. VFX. Beat. Captions.
reanimate → rebuild any animation you see, frame by frame
github.com/NiazMorshed2007/r…
motion-design-skill → timing and easing like a trained animator
github.com/LottieFiles/motio…
mg-styles-15 → 15 motion styles, one prompt each
github.com/Vincentwei1021/mg…
huashu-art-motion → paintings that move in 35 art styles
github.com/alchaincyf/huashu…
anidoodle → doodles and loops, all in code
github.com/alexgreensh/anido…
cinetic → launch films straight from code
github.com/Leonxlnx/cinetic
motionmaxxing → turns AI slop into designer-level motion
github.com/Tejashmakwana/mot…
Manim → turns formulas into smooth animation
github.com/ManimCommunity/ma…
blender-skills → pro Blender tricks your agent can use
github.com/EL4CTEO/blender-s…
three.quarks → particle effects that make every scene pop
github.com/Alchemist0823/thr…
react-postprocessing → the cinematic finish on every frame
github.com/pmndrs/react-post…
Tweakpane → sliders for every tiny detail
github.com/cocopon/tweakpane
EffectCraft → fast compositing, built in Rust
github.com/storytold/effectc…
Openverse → free sounds and music for every scene
api.openverse.org/v1/
librosa → finds the beat so your motion follows it
github.com/librosa/librosa
WhisperX → word-perfect captions, timed to the millisecond
github.com/m-bain/whisperX
Remotion Skills → official video skills for any coding agent
github.com/remotion-dev/skil…
FFmpeg → turns your animation into a real video file
github.com/FFmpeg/FFmpeg
Copy → style → move → sound → ship.
Bookmark this, then open the guide below and animate something nobody can scroll past ↓
Miguel Miranda retweeted
Oh boy … PHP sent clear text data when you used uppercased S in httpS:// in a protocol handler 🙈🙈
file_get_contents('HTTPS://…') skipped TLS: tokens, POST bodies and FTPS logins went out unencrypted!
Fixed today in 8.2.36, 8.3.37, 8.4.28, 8.5.13 & 8.6.0.
Vibe coders are getting sued.
The app you hacked together and shipped over the weekend is 𝗮𝗹𝗿𝗲𝗮𝗱𝘆 𝗯𝗿𝗲𝗮𝗸𝗶𝗻𝗴 𝟲 𝗹𝗮𝘄𝘀 without you knowing:
1. Signup doesn't ask for age. Let in one kid under 13 and that's a COPPA violation, up to $53K, counted per child.
2. Load Google Fonts from their servers and every visitor's IP gets sent over. A Munich court ordered one site to pay a visitor €100 for it under GDPR.
3. Use session replay to record what users type without consent, and in California you get sued for wiretapping under CIPA, up to $5K per violation.
4. Marketing emails with no unsubscribe link or postal address violate CAN-SPAM, up to $53K per email.
5. If renewal terms aren't right next to the subscribe button, California treats every renewal as an unconditional gift. You have to give the money back.
6. Skip the $6 DMCA agent registration with the Copyright Office and you're on the hook for whatever infringing content users upload. Willful infringement can cost up to $150K per work.
Each of these 𝘁𝗮𝗸𝗲𝘀 𝗺𝗶𝗻𝘂𝘁𝗲𝘀 𝘁𝗼 𝗳𝗶𝘅. Throw this at Claude Code before you launch:
"Check my app for: missing age gate, externally loaded fonts, session replay without consent, marketing emails without unsubscribe and address, renewal terms not beside the subscribe button and no DMCA agent. Fix what you can and list what I need to do myself."
Miguel Miranda retweeted
holy f*ck this is smart
a 16-year-old cs student in china worked out how to run 300 grok bots for under $0.02 per bot per week
the same swarm would burn $4,500 a week without it
he cut pointless status checks, stopped rereading full chat histories and killed every endless retry loop
now most bots stay idle until there's real work to do
here are the 20 github repos he used to bring the cost down:
1. markitdown > convert pdfs, office files and html into clean markdown before analysis so the bot reads the content, not the layout
github.com/microsoft/markitd…
2. llmlingua > compress long inputs before the model reads them while keeping the facts, constraints and qualifications the task depends on
github.com/microsoft/llmling…
3. firecrawl > collect web content as markdown or structured data so the workflow keeps only the fields it needs
github.com/firecrawl/firecra…
4. rtk > compress common command outputs before they enter the conversation so repetitive logs don't swallow useful context
github.com/rtk-ai/rtk
5. chonkie > split long documents into chunks so retrieval pulls the relevant passages instead of the whole file
github.com/feyninc/chonkie
6. flagembedding > rerank retrieved passages before generation - strong evidence stays, irrelevant matches leave the prompt
github.com/flagopen/flagembe…
7. mem0 > store and retrieve relevant facts across runs so each task starts with selected memory instead of the entire conversation
github.com/mem0ai/mem0
8. prefect > cache successful task results and reuse them when configured inputs match, skipping reruns of steps that already finished
github.com/prefecthq/prefect
9. gptcache > serve saved answers to similar requests instead of generating again - set matching and freshness rules around the task
github.com/zilliztech/gptcac…
10. graphiti > keep changing facts in a searchable graph and pull relevant context instead of replaying the full event history
github.com/getzep/graphiti
11. portkey gateway > cache matching api responses behind one shared gateway so connected workflows reuse answers across repeated requests
github.com/portkey-ai/gatewa…
12. hishel > add http caching and revalidation to your fetching tools so unchanged responses stay out of repeated download-and-analysis cycles
github.com/karpetrosyan/hish…
13. celery > put background jobs on a worker queue - code handles waiting and dispatch, model calls run only inside tasks that need them
github.com/celery/celery
14. requests-cache > store fetched responses with expiry rules so your scripts reuse valid data before asking the bot to look again
github.com/requests-cache/re…
15. apscheduler > run lightweight checks on a schedule in python and batch nonurgent work before passing useful updates to the agent
github.com/agronholm/apsched…
16. routellm > send simpler requests to a cheaper model and keep the stronger one for work that needs it
github.com/lm-sys/routellm
17. phoenix > trace your ai workflow, inspect repeated calls and compare changes against output quality before keeping an "optimization"
github.com/arize-ai/phoenix
18. pybreaker > pause calls to a service that keeps failing once it hits a set threshold, so one broken dependency can't loop forever
github.com/danielfm/pybreake…
19. semantic router > match requests to predefined workflows with local embeddings instead of paying a large model to pick every next step
github.com/aurelio-labs/sema…
20. litellm > track api spending and set budgets before one runaway workflow eats the whole allowance
github.com/berriai/litellm
connect the tools your workflow needs through the terminal, mcp or your own code - route api calls through the gateway you picked so its controls actually apply
start with the biggest leak: oversized inputs, repeated research, empty checks or failed operations that keep restarting
filter before reading > reuse before generating > check for changes > stop failed loops
then measure finished work against total usage, including the tools running around your bot
the full course below shows how to put the workflow together ↓
Miguel Miranda retweeted
Google acaba de matar las apps de escaneo de documentos de terceros.
El nuevo escáner nativo de Drive hace capturas multipágina en tiempo real, detecta duplicados automáticamente y tiene modo continuo.
Miguel Miranda retweeted
Apple rejected my first app twice
There’s a free GitHub repo that would’ve saved me both times
app-store-compliance:
- every common rejection reason, mapped
- checks your app against 2026 rules
- AI agent audits your code before you submit
Use it before your next submit!
Miguel Miranda retweeted
A 4.4GB uncensored AI image model that can generate and edit images for FREE 😋
Meet Qwen Image 2.1 Uncensored
> Only 4.4GB in size
> No access gate
> Free to use
> Text-to-image + image editing
> ComfyUI GGUF support
> 37K+ downloads
Try it here
huggingface.co/chatrlabs/Qwe…
A 2.8GB uncensored AI model you can run locally 😋
Meet DeepSeek V4 Pro Qwen3.5 4B Uncensored
> Only 2.8GB size
> Runs on normal PCs (llama.cpp / LM Studio)
> Almost no refusals (5 out of 100)
Free and open source
try it:huggingface.co/xtools-at/Dee…
Miguel Miranda retweeted
‼️ Breaking: Hackers Infect 178 Major Websites via Optimizely to Steal Credit Cards and Target iPhones
Details: cybersecuritynews.com/optimi…
On 16 November 2026, attackers planted malicious JavaScript inside the Optimizely Web experimentation platform. The code reached visitors of 178 organizations.
The list includes some of the busiest sites on the internet, such as Zoom, the BBC, CNN, Salesforce, GitLab, IKEA, HSBC, Equifax, Air France-KLM, Ledger and Peacock.
This was not a simple skimmer. The malware worked as a multi-purpose kit. It swapped cryptocurrency wallet addresses, stole payment card details, grabbed cookies and screenshots, and pushed the Coruna and DarkSword exploit kits to iPhone and iPad users.
#CybersecurityNews
Miguel Miranda retweeted
요즘 Claude Code로 DB까지 설계하시는 분들 많으시죠?
저도 여러 앱을 만들다 보니 DB 설계가 은근히 신경 쓰입니다. 코드는 다시 짜면 되는데 DB는 운영하다 구조를 바꾸려면 일이 꽤 커지거든요.
Claude Code용 Database Architect 에이전트가 공개됐네요.
단순히 SQL만 짜주는 게 아니라
1. 테이블 구조, PK/FK, 인덱스 설계
2. PostgreSQL, MySQL, MongoDB 등 DB 선택
3. 멀티테넌시, RLS 등 데이터 격리 설계
4. 쿼리 성능과 확장성 검토
5. 마이그레이션과 롤백 계획
설치는 쉬워요 터미널에서 한 줄이면 됩니다. (요즘 대부분이 이렇긴 하지만 ㅎ)
npx claude-code-templates@latest --agent database/database-architect
설치 후 Claude Code에서 이렇게 요청하면 됩니다.
"database-architect 에이전트로 현재 DB 구조를 분석하고, 테이블 관계와 인덱스, 성능 개선안을 검토해 줘. 변경은 하지 말고 설계안부터 보여줘."
운영 중인 DB라면 바로 수정하지 말고 검토부터 시키는 게 좋겠죠.
저도 한번 적용해 봐야겠습니다. ㅎㅎ
Miguel Miranda retweeted
BREAKING: A cyberattack hit 178 major organizations, including Zoom, CNN, BBC, Fox News, HSBC, Ledger, Equifax, Salesforce, IKEA and Air France-KLM, by loading malware on the Optimizely Web platform, exposing visitors to scripts built to steal payment data and exploit iOS devices, per security firm Report URI.
Banking and finance was the hardest-hit sector, with 32 organizations affected, including the UK's HSBC across multiple markets such as the US, Hong Kong, Mexico, Singapore, Australia and the UAE.
Miguel Miranda retweeted
This is a big one.
We found malicious JavaScript being served through Optimizely to some of the world's largest websites.
178 organisations affected, with everything from crypto theft and card skimming to iOS exploits.
Zoom, BBC, CNN, HSBC, Salesforce, GitLab, KLM, IKEA, Equifax, Ledger. The list goes on... Full list in the post.
blog.report-uri.com/worlds-l…
Miguel Miranda retweeted
I taught a neural network to drive an F1 car.
Can it beat the 2026 pole at Spa?
0 h: every car crashed.
213 h of practice: first lap, 2:03.26.
4,266 h: 1:45.35.
Pole was 1:44.361. Not yet.
Built as a test driver for my 2026 F1 car model.
#F1 #AI #ReinforcementLearning































