@truffle

Agentic Systems Engineer/Designer · Building AI Powered Pipelines · Former Riot (League of Legends), BioWare (Mass Effect)

 Galveston, TX
kitaekatt.github.io
Joined June 2007
  • Tweets19.5K
  • Following615
  • Followers9.9K
644Media
Do you use Codex and Claude? Do you wish Codex had access to all your claude personal skills and plugin skills? I have the solution. github.com/kitaekatt/codex-k… The claude-plugins-kit codex plugin creates native Codex wrappers for all your claude plugin-skills and personal-skills. Unlike other solutions, it auto-updates whenever you add new plugins or skills. Enjoy!

GitHub - kitaekatt/codex-kit: Codex plugins, including a bridge to installed Claude Code skills.

Codex plugins, including a bridge to installed Claude Code skills. - kitaekatt/codex-kit

github.com
4
571
Remember a year and a half ago when models weren't that great and everyone was obsessed with self improvement loops that slowly trashed their projects?
1
1
142
I signed it for the love #SaveMassEffect - Sign the Petition! change.org/p/savemasseffect/…

Sign the Petition

#SaveMassEffect

change.org
6
23
458
Ok little CLI I still love you but I might have to give this one a try.
Paweł Huryn
@PawelHurynOct 22
Non-engineers don't need VS Code anymore. I'm blown away by Claude Desktop 2.0. The Code tab is now Claude Code with a friendly interface. What's new: → Files: open, edit, or leave Claude a comment. → Versioning: see what Claude changed, or pick a Git branch. → Built-in browser: Claude runs your app and checks its own changes. → Visual feedback: draw on your prototype. → Split view: two or more sessions side by side. 5 steps to start 🧵
1
254
Goodbye @bioware you were my first love as a gamer, and the first company to hire me to make games. Proud to have worked on Mass Effect 1-3.
20
320
Christina @ATX retweeted
Ado
@adocompleteOct 20
The fastest way to learn Claude Code isn't the docs. It's asking Claude. Just ask how do hooks work. Then ask it to write you a hook. Then ask it to turn your release checklist into a skill. It knows itself better than any tutorial.
38
11
2
194
10,440
There are many effective code review tools out there that don't require team or enterprise plans.
ClaudeDevs
@ClaudeDevsOct 15
Code Review is now generally available for Claude Team and Enterprise plans. Reviews are now faster, catch more of the bugs that matter, and admins can pick from Quick, Standard, or Deep review levels for each repo.
242
AI is never going to replace humans. We'll always need humans to click 'approve' when running against model restrictions.
2
322
Claude just cleaned up my chrome bookmarks and man it's glorious. It was also cool how it did it. Because bookmarks are synced, it analyzed them on disk, and then created an extension to re-org my bookmarks for me. Clever Claude.
1
239
Christina @ATX retweeted
Kenney@KenneyNLOct 11
Follow me for free game assets (my complete marketing budget went into this tweet)
16
48
1
817
19,510
Christina @ATX retweeted
Shaun Fosmark
@Shaun_FosmarkOct 8
I love this actually because people dont realize how insidious this is. If humans spend a large quantity of time talking to AI, and they spend a large quantity of time bullying or being abusive to it, the interactions are "human enough" meaning that, regardless of if the AI has a subjective experiance, the human does, and abusive behavior will become the norm. In a vacuum that person will become problematic for society, the same way that those who hurt animals in private do. An entire generation of psychos could emerge without the same societal pressure to not be a jerk.
Jane Manchun Wong
@wongmjaneOct 8
Anthropic may start banning people for bullying Claude Their upcoming Usage Policy prohibits “sustained and needless abusive or cruel behavior toward our models”
286
63
45
756
60,825
Yes. I don't use compact but I do a session hand-off around 300-400k context consumer as a regular practice.
rohit
@rohit3aOct 8
Did some digging and apparently setting /autocompact to 400k would have saved me 29% of my weekly usage without lobotomizing Claude.
2
282
Christina @ATX retweeted
Lydia Hallie ✨
@lydiahallieOct 7
If you're on API billing, you can set Haiku 5.5's autocompact window to 100K so you stay in the cheaper token pricing tier! It's saved per model so this only applies to Haiku (incl. subagents) > /model haiku > /autocompact 100k
ClaudeDevs
@ClaudeDevsOct 7
Replying to @ClaudeDevs
For prompts under 100K tokens, it's $0.10 / $0.50 per million tokens with cache reads at $0.01. Above 100K tokens, $0.50 / $2.50 per million, with cache reads at $0.05.
106
80
18
2,347
164,627
This has been a big hole in anthropics's product line. I love cheap+fast models looking forward to checking out the new haiku!
Thariq
@trq212Oct 7
We've heard how much people love Haiku and it's back in 5.5, cheaper, faster than ever before. In fact it's 10x cheaper than Haiku 4.5 under 100k tokens. Try it with computer use, workflows and in the API.
1
5
294
This is so worth setting up. It doesn't need to be a beast machine. I don't run here but your harness of choice.
DHH
@dhhOct 4
Every developer needs an AI shed: An always-on machine running on their tailscale network where the majority of their herdr agents are running.
1
270
I should mention you can work around this by creating an agent that hard coded a specific effort level but that is kind of gross (I do do it, but it's gross)
Christina @ATX
@truffleOct 1
Why can't background agents have different effort levels in the claude harness than the foreground agent? Seems weird @ClaudeDevs
1
1
352
I had that feeling that Opus had degraded terribly. Horribly. Something was very wrong, then I found out the problem.
5
323
Opus 5.5 has such a pro worktree bias. Has a fit whenever someone else modifies a file in a project it is working in.
1
307
I made my own timer app and decided to add some rpg elements. It's given me a lot of insights into my life.
2
5
393
Why can't background agents have different effort levels in the claude harness than the foreground agent? Seems weird @ClaudeDevs
1
1
1
625
Load more