|
[AI Engineering]
Dynamic Workflows in Claude Code
Anthropic has rolled out dynamic workflows natively within Claude Code. The agent can now map out and coordinate dozens to hundreds of parallel subagents to tackle heavy engineering tasks independently, benchmarking and checking its own code adjustments before delivering the final fix to your terminal.
Explore the architecture →
|
|
[Productivity]
Natively review PRs inside Linear
Context-switching fatigue is real. Linear just launched Linear Diffs, allowing engineers to view, manage, and audit full code branches and pull request diffs directly inside your issue tracking view. No more jumping back and forth to external Git layouts just to sign off a ticket.
Read the changelog →
|
|
[AI Regulation]
US Directive restricts Fable 5 and Mythos access
A major shift in geopolitical tech access: the US government has issued an export control directive ordering Anthropic to suspend foreign nationals' access to their highly advanced Fable 5 and Mythos 5 frameworks, signalling tighter barriers on foundational high-compute AI models.
Read the official statement →
|
|
[Project Boards]
Markplane
An AI-native, markdown-first management system. It builds project trackers and lanes natively within your code repository structure.
Github Repo
|
|
|
[Mocking]
Robohash
An incredibly fast web endpoint to generate random, unique, and clean avatar graphics directly from any text strings. Great for seeding databases.
Visit Tool
|
|
|
[Internal Knowledge]
Hacking Claude Code session limits & verification
• Juan Salvador: If you use Claude Code heavily, don't wait to hit the afternoon token wall. You can trick the session window by scheduling an automated daily Claude Code Routine 5 hours before your day begins (e.g. at 7:30 AM). Have it use Haiku to simply say "hello". This kicks off the master session loop early, moving your midday usage reset back to 12:30 PM instead of forcing you to stall later. Check the panel at claude.ai/code/routines.
• Compliance Alert: Anthropic is introducing mandatory Identity Verification rules starting July 8, 2026, to access specific high-tier operational workflows. Make sure to complete your profile ahead of the deadline to prevent runtime workflow lockouts.
Read Anthropic's ID notice →
|
|
|
MarsBased updates
|
10K+
|
Community
10,000+ podcast subscribers!
Our community keeps hitting new milestones at record speed. We have just surpassed 10,000 subscribers on the MarsBased Podcast channel. Thank you all for listening, sharing our shorts, and being an essential part of this journey!
Enter the community →
|
|
Podcast
"I don't want to hear that Claude told me...": Road to CTO with Xavier Tous
Àlex Rodríguez Bacardit sits down with Xavier Tous to discuss the reality of tech leadership when AI becomes a daily tool. Moving past the usual hype, they dive into why engineering judgment matters more than ever and how to prevent teams from delegating critical analysis to LLMs.
Watch the episode →
|
|
Article
Using AI to run performance reviews: How we evaluate our team with Claude
We open up our internal playbooks to show exactly how we leverage Anthropic's Claude to parse, structure, and eliminate bias from our team's quarterly evaluations seamlessly, keeping human craft at the core of our loops.
Read the post →
|
|
Team
Welcome Javier Párraga, Full Stack Engineer!
Javier is a Full Stack Engineer who's been building his engineering career across diverse tech companies. Outside of work, you can find him at the gym or locked into a video game. He is currently chasing a milestone and is close to hitting a 140kg bench press!
Meet the Martians →
|
|
