ทวีตที่ปักหมุด
beyondr🏴☠️𓂀
1.5K posts
beyondr🏴☠️𓂀
@thebeyondr
building tools with some ai features
embed space เข้าร่วม Ağustos 2020
1.5K กำลังติดตาม277 ผู้ติดตาม
actually having fun making this harness ruleset by figuring out the mental model. now to turn it into proper agent harnesses and commands and test in a project! @zakelfassi any tips on (agentic coding) harness testing?
English
biggest learnings so far:
+ task queueing via celery + redis
+ multi tenant considerations (ensuring data for one user doesnt leak in another user's request response)
+ @pydantic 's neat `pydantic_ai` package for model interactions
English
the stackeroo:
django + celery + react, gpt-4.1 nano for classification.
next up: theme discovery and the self-improvement loop. 🫡
English
currently working on a system that takes your customer's feedback and gives you actionable insight.
ingestion pipeline and AI classification are working! sentiment, themes, urgency all auto-tagged.
you can also correct anything the AI got wrong and it learns from that.
English
beyondr🏴☠️𓂀 รีทวีตแล้ว
@cursor_ai @ryolu_ this is annoying because it replaces the pervious response completely in the sidebar
> command agent
> agent does work relate to command
> shifts all content up
> activates its own command
English
no UI (for now?) but its a multi-agent FastAPI pipeline: parallel collectors, pgvector dedup, LLM synthesis, retries, dead letters, structured logging, then it renders an executive HTML brief and sends it on a schedule.
Set targets, walk away, get the brief Monday morning.
English
it ain't perfect but it lives.
Autonomous competitive intelligence platform for continuous monitoring, analysis, and scheduled executive brief delivery to your inbox
github.com/unified-method…
English
learning a lot building a system that monitors competitors automatically. checks websites for price changes, scans news, watches social media. one weekly email with what changed and why it matters. if one source goes down you still get a report from the others
English
@nikitabier @benjitaylor I hit message icon from someone's profile who I've never spoken to, I'm prompted for 4 digit code, the chat interface forgets who I was trying to start a chat with. have to go back and hit the message button.
English
Calling all talented software engineers: Varick is hiring.
But we value your time so the entire application to interview process will last < 3 days. And if you refer someone we hire, we'll send you $5k.
@varickai deploys production grade AI systems inside large enterprises that actually realize ROI. We're actively working with massive companies and are drowning in demand. We're a small and talented team in SF, moving to a beautiful office.
You get to:
- build agents all day with 0 politics
- leverage an unlimited token budget
- access entire departments of billion dollar enterprises
and every piece of your work results in $10Ms of enterprise value.
We'll give you very good compensation, and you're early enough that your work shapes everything.
Relevant experience is a hard requirement.
If you're interested send a resume + short blurb (name, employment history, university) to contact@varickagents.com, and I'll book 20 minutes to chat and see if this is a good fit. Engagement greatly appreciated to boost this around!
English
Giving away 5 Opencode Go subs
Winners selected randomly from comments in 24 hours.
OpenCode@opencode
we’ve signed Zero Data Retention agreements with all providers for Go all models now follow a zero-retention policy your data is not used for training
English
"serious" is just seasoning in this tweet
kache@yacineMTB
If you are doing anything serious you very quickly see the limits of these new AI models
English
he's not wrong but I think most will skip the fact that he's talking from his perspective. its like taking advice from rich people that you should chill and money isn't everything.
I understand what phase I'm in. I'm not there yet.
dax@thdxr
right now everything in the world is telling you to go faster, ship more, add that feature, start another project so i'm actively working on feeling ok not doing any of that
English
how I'm learning: x.com/thebeyondr/sta…
beyondr🏴☠️𓂀@thebeyondr
I make a meta-[project-name] folder. Subfolders are: colearn, resources, and the project folder. colearn folder has the curriculum, teaching criteria and my workbooks. resources are for project scope and plans, maybe docs. actual project is built in phases after checkpoints
English
colearn/ folder now has: learning curriculum, grading criteria, milestone workbooks with my predictions and grades.
for me rn, code is half the battle. the other half is proving I understand the decisions being made
English
the actual build cycle: i predict + narrate how id build it. claude grades against a criteria doc. i retry if i want a better grade. then it builds. then it writes build notes: what matched my predict, what didnt, and what file:lines to study
English
I make a meta-[project-name] folder. Subfolders are: colearn, resources, and the project folder.
colearn folder has the curriculum, teaching criteria and my workbooks.
resources are for project scope and plans, maybe docs.
actual project is built in phases after checkpoints
0xSero@0xSero
How I learn these days, maybe this will truly hollow my brain out, who knows.
English
"the only true test of intelligence is if you get what you want out of life"
— @naval
Prompter@PromptLLM
your output from AI is a direct correlation to your IQ
English
Been trying the 45/15 split for like 2-4 rounds early in the day recently and I like it a lot. Phone in another room when working. Guilt free 15m after. Usually get a shadow box round in then goof off for the rest. I get way more done vs trying to be productive all day.
EP@eptwts
my best work is done when i open my laptop with a specific task in mind, do 2-4 hours of work & then close my laptop for the rest of the day... my least productive days are the ones where i'm on my laptop for 12 hours clicking between tabs & doing random low ROI tasks
English