Brad Abrams retweeted
⛑️ safety tip for Claude Managed Agents
Run /claude-api managed-agents-secure in Claude Code to compare your agent's settings with the security guide. It's read only and lists each setting to change.
platform.claude.com/docs/en/…
Brad Abrams retweeted
We've raised the max request size for Messages and Token Counting on the Claude API from 32 MB to 128 MB.
Image-heavy requests like long computer-use sessions and big PDFs were previously hitting 32 MB with plenty of context window left. Now you get ~4x more room.
Brad Abrams retweeted
Brains <> Hands is the way to go for so many reasons. The upsides to reliability, security, latency, and scalability far outweigh any perceived downsides, which are extremely solvable with just good engineering practices (or just telling Claude "pls fix")
Super happy to have played a smol role in popularizing this architecture alongside @gcemaj.
How Anthropic ships Claude with Claude
latent.space/p/opus55/
Opus 5.5 is the #1 LLM in the world right now. We go behind the scenes of the launch, with @katelyn_lesse and @angjiang, heads of Eng and Product behind the platform used to build everything from the first party harnesses like Code/Tag/Duet and by >50% of Anthropic's top customers!
Trust your agent. But read its logs.
A few resources to help if your agent is on Claude Managed Agents:
Read the guide: platform.claude.com/docs/en/…
Run in Claude Code: /claude-api managed-agents-secure
Review its sessions: platform.claude.com/cookbook…
Brad Abrams retweeted
We’ve added a new way to access Claude on Google Cloud.
You can now use the full Claude API from your Google Cloud account, including Claude Managed Agents, code execution, Opus 5.5 fast mode, and new features as they launch.
Brad Abrams retweeted
Seeing lots of struggles with API keys.
Started using WIF in more places recently so wanted to share in case this helps: platform.claude.com/docs/en/…
Do you know what your agents did last night?
New cookbook: point Claude at your agent transcripts and it flags the sessions worth a second look. Works for CMA, Agent SDK and your own agent transcripts
platform.claude.com/cookbook…
I built a set of strategies with dynamic workflow for Claude Managed Agents: Advise, Grade, Fan-out, dream.
linkedin.com/pulse/give-your…
What other patterns would you like too see?
Building secure cloud-hosted agents for models this smart and persistent is not a weekend project.
We’ve put that work into Claude Managed Agents. For example, your API keys stay in a vault and the sandbox only ever sees a placeholder.
We just published a guide to locking down a deployment. platform.claude.com/docs/en/…
How confident are you in your agent sandbox?
Brad Abrams retweeted
Fantastic writeup here on Flows and why we're all so excited about it:
latent.space/p/claude-dynami…
Brad Abrams retweeted
You can kick the tires with Claude Managed Agent Duynamic workflows, today!
Checkout the Claude Console (platform.claude.com) and try out one of the templates with "Dynamic workflow" in the card.
world’s first full length workshop on claude flows
@cjav_dev and Harrison Stall are kicking off @aidotengineer bright and early on a monday morning to a PACKED room, standing room only #AIEngineerNY
@ricmac is on the case reporting for @latentspacepod and we’ll also publish asap. proud to work with @anthropicai to sponsor AIE NYC for the second year running!
Brad Abrams retweeted
big morning in NYC today talking about flows! what a blast, pumped for everyone to try things out. let us know what wild problems you’re able to solve with the platform!
world’s first full length workshop on claude flows
@cjav_dev and Harrison Stall are kicking off @aidotengineer bright and early on a monday morning to a PACKED room, standing room only #AIEngineerNY
@ricmac is on the case reporting for @latentspacepod and we’ll also publish asap. proud to work with @anthropicai to sponsor AIE NYC for the second year running!
Brad Abrams retweeted
I pointed the new Claude Managed Agents dynamic workflows at 18k lines of code.
> 21 proven bugs.
> 5 minutes.
> $8.36, on my Claude Max API credits.
A lead agent writes a plan, runs it across many agents in phases, then merges the results.
Try it in Claude Code:
/claude-api managed-agents-onboard bug-hunter
It burns tokens fast. It's worth it.
Brad Abrams retweeted
This was a labor of love from so many people and I'm so excited to see it out for everyone.
Dynamic Workflows are one of the most powerful capabilities you can utilize for uplift on your agentic workloads. Time and time again we've seen that letting Claude stitch together agents, context, and data programmatically leads to better outcomes. It also keeps your "Big Claude" free to talk to the user and orchestrate "lil Claudes", instead of spending time micromanaging individual threads.
This first release is only the beginning for Dynamic Workflows in CMA. We're going to invest a ton in giving developers more control and capabilities when working with workflows, and generally making the harness even more powerful, reliable, and flexible.
Massive shoutout to @karbon0x who was the mastermind behind Workflows for Claude Managed Agents.
Have fun Flowin' ya'll. Let us know what you think!
This was such fun to work on! Turn it on with one field on your agent, then ask for a workflow in a new session.
Tip 5: Two quick ways to start:
In the Console, pick a quickstart template and try it in a test session.
In Claude Code, type /claude-api managed-agents-onboard and it walks you through the setup.
Docs: platform.claude.com/docs/en/…
Can't wait to see what you build!
Tip 4: Make it better. Steer the workflow in plain words:
"Have a second checker review every scene before the final cut."
Every scene, not only the flagged ones. If one fails, redo it and check again.
More checks use more tokens, so spend them where it matters.
Tip 3: Curious what your agent built? Ask it. I asked mine to show me the workflow it wrote, and to draw the run.
"Print the workflow that you started the run with, word for word, in one code block. Do not start another run. Draw the workflow run as a diagram and save it to workflow-diagram.png, 1600x900, light background: phases left to right, one box per agent, arrows for what feeds what. Do it yourself; start no workflow run."
