Posted by datadrivenangel 4 days ago
They don't even allow alternative harnesses on their subscription model. (But they do allow a agent-sdk that can pretend to be a generic compatible endpoint the alternative harnesses can use).
If you like your stuff overengineered, gated, and over-priced then by all means.
Its not even wrong from Anthropic because it's the path to most profit.
I'll take the opinion of people who have strong opinions about mild inconveniences.
Between my goals and Anthropic's bottom line, i want my goal to win out without needless inconveniences.
If you think I'm wrong, and these arguments have nothing to do with it, just tell us what explains Claude not supporting AGENTS.md.
Was it such a grand engineering challenge? Too many 'legacy' projects relying on both AGENTS.md/CLAUDE.md and that has recently changed?
I've made clear my belief on the reason behind it - branding decisions outweighing utility.
Add something to the thread by sharing your version of a 'more likely' decision process that led to the choice to only now support it.
> Our reasoning for this was that we don't think that model families are interchangeable and the system prompt can have a big impact on performance.
I don’t disagree with anything you say, I just don’t think AGENTS.md ever was a big deal, just a small annoyance
Sometimes, Anthropic has pushed standards (MCP, Skills), and that is great, but other times they have explicitly built incompatibilities. With the Plug-in marketplaces, it is quite similar.
When prompted to write to it, it will always try to write into the Claude.md and error out with that approach.. It eventually figures it out, but it's a ymmv situation I think
How is agents.md and different? Just a standard for other models? Is claude the lower standard now?
Sent from my iPhone.
I was quite annoyed with Claude earlier this year when I had to maintain duplicate files or rely on symlinks.
Seemed like a waste of time simply because Anthropic didn’t want to take the “L” and accept that AGENTS.md was the new standard.
IMO, this kind of product decision making is why so many products either suck, or they’re are good but not great … and users have zero loyalty.
I find it super annoying this building lock in instead of loyalty approach to product seems to be standard practice these days.
404
I didn't even have that symlink in any other project - it just did it. I think it saw that one of the projects I already had was set up by Codex and that project had an AGENTS.md so perhaps it inferred that I was using both Claude and Codex, so it was politely covering both? Or maybe a recent change made this behavior default?
I was surprised and I hope they continue to seek standards.
Oh-my-py unlocks max capabilities at the smartest token usage, and you can change AI company and providers while still getting things done. Roles, advisor, and the ability to use Chinese models on international providers with no training and ZDR behind OpenRouter are very powerful. You can have providers and model companies competing on price, intelligence, and speed for something that's useful but with less and less moat as model progress plateaus.
I now use GLM 5.3 for slow/planning, GLM 5.3 Flash as the default task-vision model, and DeepSeek Flash 4.1 as the advisor. Providers are ordered by cheapest at the minimum TPS I want for that model. I use it not only for coding but also as a personal assistant. I can ask it things and it gets them done automatically, like opening an issue on a repo, then having another agent ship the code, run the tests, and update master after having changed the app to autoupdate on cut release.
I can also ask it to do groceries. I use Bitwarden machine secrets and a limited virtual card connected to my phone.Something else that's very powerful is asking it to launch parallel agents to do things you know don't overlap too much.
The amount of agency we're gifted with is crazy, you can now move hundreds of hours of work with one hour of input using an open-source harness and open-weight models. I can't wait for more competitors in the chip space, and to be able to buy models etched on silicon for even higher agency, speed, and independence.
I also use the same. But I use Anthropic models (Opus 5) with it.
Maybe internally they see really unbelievable things, but my impression is that they pushed so hard on agents that they don't have the grasp of the situation.
They win either way
mkdir -p ~/.githooks
git config --global core.hooksPath ~/.githooks
cat > ~/.githooks/post-checkout <<'EOF'
#!/usr/bin/env bash
if [ -d .agents/skills ] && [ ! -e .claude/skills ]; then
mkdir -p .claude
ln -s ../.agents/skills .claude/skills
fi
EOF
chmod +x ~/.githooks/post-checkoutDoesn't look like Anthropic care about dev community
Shocker.
In the manner of someone finding a dead mouse and holding it up for examination CC said it could find no instructions but perhaps it should check this AGENTS.md file.
Very sassy, Codex!
Also: did it suggest instructions for the “correct” agent, or an ambiguous agent?
One has marketing implications, the other one is generally decent advice.
> The function you added is load-bear-very important [...]
I never felt this mocked by a computer.
IMPORTANT: this context may or may not be relevant to your tasks. You should not respond to this context unless it is highly relevant to your task.
IMPORTANT: this context may or may not be relevant to your tasks. You should not respond to this context unless it is highly relevant to your task.
Codex follows the AGENTS.md far better. CC seems to have nudged people away from taking the CLAUDE.md as mandatory instructions.
This bug was closed Not Planned and from a recent analysis of the system prompt the behavior is still there even with the new inclusion of AGENTS.md.
But this is a good change.
The world of managing skills between providers still feels messy though.
https://code.claude.com/docs/en/memory#import-additional-fil...