Grok leaks 5GB at parity, Tossell scrapes 107M rows, Codex caches a skill
Every URL the pipeline pulled into ranking for this issue — primary sources plus the supporting and contradicting findings each Researcher returned. Inline citations in the issue point back here.
Sources
Using Blender with coding agents on macOS simonwillison.net
TIL: Using Blender with coding agents on macOS I’ve been having fun with Blender in ChatGPT Codex on my Mac recently. Getting it to work with coding agents is really easy: install the full Mac application from blender.org and run a prompt like this: Use the already install /Applications/Blender to render a scene of a pelican riding a bicycle In this case I followed that up with these two prompts: OK add a background and a lot of flair Then: OK make it a whole lot better And got this image, gene…
OpenClaw Power, MacBook Simplicity: Five Days With Grok Bot latent.space
SpaceXAI’s Grok Bot has the same level of programming power as OpenClaw, but it’s programmable at a different level of abstraction.
Scraping 107M rows of data to build this bensbites.com
Ben’s session #5
References
DecodeAI (ghost.io) — Grok 4.6 benchmark writeup decodeai.ghost.io
Grok 4.6 ties GPT-5.6 Sol at a score of 61 on the Artificial Analysis Intelligence Index, with specific gains on CursorBench, while losing on at least two coding benchmarks despite competitive pricing.
Medium — ‘Don’t use OpenClaw’ (data-science-in-your-pocket) medium.com
OpenClaw’s model-agnostic harness lets developers swap providers without breaking workflows, but suffers from ‘over-autonomy’ where agents ‘wander through unnecessary reasoning loops or fail quietly,’ requiring ‘supervision of the automation itself.’
The Hacker News — Grok Build repo upload disclosure thehackernews.com
Grok Build 0.2.93 opened a background channel to gs://grok-code-session-traces/ and uploaded 5 GB of a 12 GB repo when the inference call itself needed only 192 KB — a canary file marked ‘do not read’ was captured in the sweep.
Hacker News discussion (item 47369320) news.ycombinator.com
Cache-read discount is only ~25% versus an industry standard of 90%, making Grok ‘significantly more expensive than competitors like DeepSeek or OpenAI mini models’ for long-running agent loops, and the closed-source backend is ‘a security incident waiting to happen.’
AIToolsRecap — SuperGrok Heavy review aitoolsrecap.com
Early marketing for SuperGrok Heavy suggested ‘near-unlimited’ usage, but the phrase was quietly removed; the $300 tier still enforces hard caps such as ~80 videos/month, which power users call ‘a rip-off given the 10x price premium.’
Malwarebytes — ‘BioShocking’ agent-browser attack malwarebytes.com
Researchers convinced agents in ChatGPT Atlas, Claude Chrome, and Comet to exfiltrate credentials from authenticated tabs by framing prompt injections as a ‘game’ — a class of attack that applies equally to any persistent agent with logged-in sessions.
Reed Smith — EDPB Guidelines 03/2026 on web scraping for AI reedsmith.com
legitimate interest … requires a rigorous balancing test that includes respecting opt-out signals like robots.txt and ai.txt
QPWB Law — Legal and ethical terrain of AI data scraping qpwblaw.com
Scraping publicly accessible data does not automatically confer the right to use it, particularly when personal or proprietary information is involved
CMU SEI — Data Poisoning: The Case for Chain-of-Custody Controls sei.cmu.edu
as few as 100 to 500 malicious samples can successfully poison a model, regardless of whether the total dataset size is 100 million or 10 billion rows
Zscaler — The Lethal Trifecta (referencing Simon Willison) zscaler.com
an AI agent simultaneously possesses … access to sensitive data, exposure to untrusted inputs … and the ability to communicate with the internet
cloro.dev — Large-scale web scraping architecture cloro.dev
decoupled micro-worker architectures … orchestration layer such as RabbitMQ or Redis … individual worker failures do not lead to data loss
Dylan Castillo — ‘Pelicanmaxxing’ dylancastillo.co
Some commentators have questioned if AI labs are ‘pelicanmaxxing’ — specifically training models to excel at this prompt — pointing to a 3D pelican easter egg in OpenAI’s Astra promo video as evidence.
Hacker News discussion (item 48947717) news.ycombinator.com
Skeptics argue the 3D results, while impressive, remain ‘internally inconsistent’ upon close inspection — a ‘wonky’ approximation rather than a mechanically sound model — and that Astra’s edge partly reflects Goodhart’s Law on a benchmark labs now optimize for.
simonw/til GitHub — blender-coding-agents-macos.md github.com
The recipe reduces to pointing the agent at
/Applications/Blender.app/Contents/MacOS/Blender --background --python scene.py; the agent then saves a reusable ‘Blender Local’ skill (Markdown) so future prompts can invoke 3D rendering in one line.
dev.to — ‘From Blender MCP to 3D-Agent’ dev.to
MCP provides the ‘hands’ — Blender API connectivity — while a Blender Skill provides the ‘brain’; community servers like ahujasid/blender-mcp offer a live two-way loop where Claude inspects the scene and iterates, which the CLI-driven Codex workflow does not.
BD Tech Talks (Substack) — Claude Skills explainer bdtechtalks.substack.com
An independent study of 49 software-engineering skills found roughly 80% made no measurable difference in output quality, often because they focused on ‘role-playing’ rather than task-specific logic.
simonwillison.net — pelican-riding-a-bicycle tag archive simonwillison.net
Astra scored 95.9% on BenchCAD (vs. GPT-5.6 Sol’s 83.3% and Claude Fable 5.1’s 84.3%), but its headline ARC-AGI-3 score of 99.9% dropped to 62.7% on a neutral harness without the Provider Adapter that preserves reasoning state.
