Leonid Bugaev @buger

Building Proof to reduce the work needed to trust a software change. Engineer, founder. I want fewer compromises between shipping and quality. reqproof.com Istanbul Joined October 2007

Tweets

4K
Followers

6K
Following

6K
Likes

7K

Leonid Bugaev @buger

3 hours ago

Those who updated to a new SuperGrok billing - does the cursor and grok bot now a unified token counter ? One of the benefits of the old plan was that you essentially got 2 subscriptions for the price of one - Cursor and Grok one. So I wonder if I should delay upgrade.

0 0 1 113 0

View Details

Leonid Bugaev @buger

3 hours ago

@boringdev77 grok bot

0 0 1 14 0

View Details

Leonid Bugaev @buger

4 hours ago

While we have so many models out of there, only a few popup up - which are smart enough for agentic flows, less then $0.5 per output, and have providers with zero retention policies. Literally just 2 models so far. And openrouter is amazing for doing this kind of research.

0 0 1 60 0

View Details

Leonid Bugaev @buger

4 hours ago

9/9 Start with one important user job. Write a few ways a user might ask for it. Run them. Read the tool calls. Fix the first gap and rerun. Full research, examples and agent-by-agent discovery paths: spiritroom.app/blog/optimise-…

0 0 0 15 0

View Details

Leonid Bugaev @buger

4 hours ago

8/9 When a job fails, find the broken step: Could the agent find the tool? Did it choose the right arguments? Did it understand the result? Did it finish the user's job? A keyword change won't fix a misleading status label. Match the fix to the evidence.

2 0 0 13 0

View Details

Leonid Bugaev @buger

4 hours ago

1/9 Your MCP can return correct data and still fail the user. We saw it with a server exposing 99 tools: missed discovery, misread results and long recovery paths. Here are 5 ways to help agents find your tools and finish the job.

1 0 0 89 0

View Details

Leonid Bugaev @buger

7 hours ago

Link to event: meetup.com/techexcellence…

0 0 0 29 0

View Details

Leonid Bugaev @buger

7 hours ago

Going to a be a speaker at Tech Excellence online meetup next week! In this session we going to answer on main question of software engineering - how much tests is enough to prove that given feature or bug-fix is working. We prove it mathematically and then turn it to deterministic guardrails which agent can't change, and has to follow. Link below👇

1 0 0 112 0

View Details

Leonid Bugaev @buger

a day ago

@BullForceOne I'm not canceling it yet, just because of quality, and that it does not reject working on security issues, but defo it went to the bottom of my usage, simply because can't use it daily.

1 0 1 16 0

View Details

Leonid Bugaev @buger

a day ago

Kimi so far has the worst subscription offer from the major labs. Its $200 plan not only gets consumed the fastest, it also has the monthly limits. As a result I have consumed the full subscription in less then 3 days, and for the rest of the month it is useless.

2 0 1 383 0

View Details

Leonid Bugaev @buger

a day ago

@LauraLoumd 👋

0 0 0 0 0

View Details

Leonid Bugaev @buger

2 days ago

I built a Hyperbolic Time Chamber for MCP servers. It's called Spirit Room. Bring your tools and the jobs users need done. Run them across agent apps. Find where discovery, tool calls or recovery break. Fix. Repeat. Even a Super Saiyan needs the right keywords.

3 2 1 230 0

View Details

Leonid Bugaev @buger

a day ago

Your best MCP tool is useless if the agent can't find it. We tested a server with 99 tools. Agents could still miss the right one. Start with the words users use. Then inspect the words the agent searches for. That's where optimization starts. spiritroom.app/blog/optimise-…