Opaline

Nice launch. Question from how we run agents on our team: a growing share of our usage isn't interactive, it's headless runs in CI (claude -p, codex exec) kicked off on every PR. Does Opaline ingest those non-interactive sessions too, and how do you attribute them? A CI runner isn't a team member, but that's often where a big chunk of the token spend actually goes.

the "catch every frustrated message" framing is the part I'd push back on. once engineers know a dashboard is surfacing their raw messages to a team lead, including the ones where they're venting at the model, people stop venting honestly and start writing for the audience instead - which quietly kills the exact signal you're trying to capture. token cost and skill usage feel like the genuinely useful team-level metrics here, the message-content angle feels like it's optimizing for a screenshot-worthy launch post more than for something a team actually wants turned on

Does this support Claude Chat / Cowork ?

as Evren mentioned, the origins of this was as an internal tool. We identified quite some issues in our skill usage across the team, to really improve and streamline the output across the team with agents.

Skill usage is the metric I'd look at first. I have a lot of skills in my setup and I honestly don't know which ones actually fire and which ones just sit there. Can Opaline show skills that were loaded but never used?

I build analytics myself and the setup looks really smooth. One question: How do you handle model updates? A lot of these numbers probably change with every new model, so comparing this month to last month might just show the new model instead of the team getting better. Is it possible to split the metrics by model version?

Curious how you handle sensitive stuff in the sessions though Coding sessions can have a lot of things you probably don’t want sitting in a team dashboard.

I could see this being handy for teams trying to figure out where they’re actually spending time with Claude/Codex.

This is kinda the part that stood out to me So much of what happens in these sessions just stays on someone’s machine.

The per message cost thing is actually pretty interesting Seeing the total spend is one thing but knowing which messages are eating it up is way more useful.

Tested it and actually gave me some interesting insights

Finally! We're duplicating so much with no way to figure out what's been solved already

Dude this i literally EXACTLY what i needed.

I would love to learn how many of our sessions are actually producing something good, something actually valuable. What are some ideas for measuring this?

hi product hunt! evren here, builder of opaline.we’re building team-wide, message-level analytics for your claude code and codex sessions:a way to pull back the curtain on how your team actually codes with agents.why? the tools people reach for today, like ccusage or /insights inside claude code, are single-user, single-model, and pretty shallow. they’ll tell you how many tokens you burned and what it cost, maybe a high-level insight or two. that’s fine for a solo dev. but real teams rack up thousands of sessions across dozens of repos and models, and right now there’s just no good way to see any of it together.we want to change that. understanding how your team works with coding agents should be effortless. like you’d never run a website without analytics... it's obvious that you need them for your agentic coding sessions too. what’s our backstory? we first built opaline for our own team at obsessiondb (yc f26) as a little side project called “rudel,” and now i’m here to give it a proper life. we were all deep in cc and codex every day, but nobody could see how anyone else was working. when someone struggled with a skill, it stayed on their machine. when someone got stuck, it stayed on their machine. when claude told "you're absolutely right!" for the third time it stayed on their machine.(if the machine wasn't punched into already)the usage dashboards out there only showed us our own numbers, and none of them kept the history we needed to actually learn over time.so we built the thing we wished existed: a shared place to see, replay, and learn from our sessions.how is your team using coding agents today? what can’t you see that you wish you could? would love to hear from you. best,evren