Diego R. Antunes retweeted
Most LLMs get "context rot": recall decreases as context window approaches the context limit.
We wondered: do the new "System 1" decision models get context rot? So we tested Jev, Clef, and Luna Decisions across 400 questions.
Turns out decision models get context rot, too.
Still qualifying leads manually? π
Let an AI agent do the heavy lifting! @mastra just dropped a new template that makes automating lead qualification in @attio a breeze. Check it out! π
mastra.ai/templates/lead-quaβ¦
Diego R. Antunes retweeted
Weβve worked with hundreds of teams shipping Mastra agents into production. Now, it's your turn.
We're introducing Mastra Launch and offering hands-on FDE support to get your agent live within 90 days.
Let's chat: calendly.com/mastra-team/lauβ¦
Diego R. Antunes retweeted
Today we're excited to launch trace imports!
Export traces from Langfuse to the Mastra platform:
What if you could turn your files from S3, Google Drive, local storage and others into a searchable knowledge base powered by vector search, BM25 and AI agents?
You can build it in no time with @mastra Organization Intelligence template mastra.ai/templates/organizaβ¦
Diego R. Antunes retweeted
Join Charlie and I tomorrow for a workshop on how to build a company brain with Mastra Connect!
You'll learn to:
- Connect an agent to your knowledge bases
- Give it tools to look things up and take action
- Put it in your team's Slack π
luma.com/mastra-304z
Diego R. Antunes retweeted
Today we're excited to launch metrics API!
Query agent costs, latency, and errors, then use the data in your reports / apps:
Diego R. Antunes retweeted
Incredibly stoked to launch Thread View for traces in the Mastra studio.
Zoom in to see a single chat turn and its associated span.
Zoom out to see the whole conversation and underlying trace:
Diego R. Antunes retweeted
Over the last year, we've met some incredible folks building agents for clients.
So today, we're launching the Mastra agency program:
A small list of agencies certified by our FDEs, with a direct line to us and visibility into our roadmap.
mastra.ai/blog/mastra-agencyβ¦
Have you heard of Generative UI (GenUI)?
Iβve been playing around with it this week, building a data exploration agent with @Mastra, @CopilotKit and React.
Depending on what you ask, the agent dynamically generates charts, tables, or interactive cards.
mastra.ai/templates/genui-daβ¦
Diego R. Antunes retweeted
Workflow + agent schedules are now GA!
Since we launched the July beta we:
- simplified running schedules on serverless / multi-server setups
- added the ability to declare a cron schedule in an agent's file
We shipped a new template for monitoring competitor pages.
A page edit doesn't always mean an important update. Use Jev to evaluate each change, and decide if it matters to you.
mastra.ai/templates/competitβ¦
Diego R. Antunes retweeted
We're excited to launch agent signals in GA today!
Since our June alpha, we added:
- canceling queued messages
- one-off "transient" signals
- ordering guarantees
π¨We shipped a new context engineering primitive @mastra today, Agent Signals! (alpha)
πUnlocks: multiplayer steering, dynamic cacheable system prompts, agent notifications (gh/email/etc), and automatic behavior guidance π§΅
Mastra + Jev? YES! π
Just launched Competitor Monitor with Jev, a new @mastra template! Track changes to products, frameworks & docs, classify them with Jev (@typesafeai), and enable daily checks and reports.
Try it π
mastra.ai/templates/competitβ¦
Diego R. Antunes retweeted
Build agents
Make money
Build Personal Agents Hack Β· Sun, Oct 4 Β· San Francisco
Total prize pool: $279,500
+ $1,625 in credits for every hacker
RSVP: luma.com/build-agents
Diego R. Antunes retweeted
Today we shipped our #1 feature request for Mastra Factory: it now connects to Jira, GitLab and Incident.io
Connect your backlog and start churning through tickets:
Diego R. Antunes retweeted
We have an intuitive understanding of how auth for humans works. But what about auth for agents?
On Thursday, @bookercodes is walking through things like resource-level permission checks, multi-tenant access, and fine-grained context for agents:
luma.com/mastra-xy3o
Diego R. Antunes retweeted
Here's the blog post:
mastra.ai/blog/introducing-mβ¦
Diego R. Antunes retweeted
Today we're launching memory hooks!
Lifecycle hooks (πππΎπππππππππππππππ, πππππππππππππ΄ππ, etc) to trigger external events
Transform hooks (πππππππΎππππππππππ, πππππππππππππππ, etc) to alter how memory is stored:







