standup-opus

# Transcript: 2026-05-05

> 1 time blocks from 8:01 AM to 8:10 AM

---

### Team standup progress update
**8:01 AM - 8:10 AM PDT** | *meeting*

**Microphone:**
We can pick some good — all of the indexing, default enrichment, analytics jobs all seem to be working really quickly, including system people. And I'm going to see how that goes. Kind of love it. Yeah, so thanks for all that.

Okay, why don't you kick us off?

Yeah, yesterday was wrapping up some final changes on Rally — guess SmallWorld nomenclature — and how the LLM handles that. Functionally, the suggested prompts are working exactly as I think we'd hope. Ran through additional testing alongside the changes that have been made. Once it gets approved, I'll probably do one quick round before getting that out. I know David really wants to show it at the Postman call this afternoon. Beyond that, as Michael mentioned, got in that dirty set fix for the indexing job, and technically testing went fairly well, presumably there. Yeah, nothing remarkable to note. And then it depends on both SmallWorld Rails side changes — yeah, we'll obviously jump on any PRs I get open today and hopefully begin breaking ground on some of that work. But that's me.

How about you, Todd?

Yeah, a tiny ticket. Yeah, I fought all day with it. I'm hitting my "can't delete anything" issue once again. I told Claude, quit deleting files. And so the new code uses the new single file configuration, but hopefully it won't be difficult to mirror this again. Test coverage is still at 87%, so it's still good for packaging. I'm hitting my Cloudflare Tunnel again properly, which is good. I think all the configuration is having the dev file and the dev setup script telling it not to suggest deleting files and using anonymous connection requests. I'm just like, stop. So now it's in the CLAUDE.md file. So hopefully those — the use of those other SmallWorld server instance underscores, I put a comment in that they're not to be used. Since they can't delete, they'll be there, but then I'll be able to deprecate the code from the back sync, which will never be able to be deleted. But even those deletes — the implementation of the functions can be, I don't know, using a nice setup for hopefully future use to get rid of the old stuff. I don't know, this probably won't require a release or anything, but I may just push a PR up there. I don't know if we're gonna do like a 2.81 dev branch, maybe perhaps — that kind of worked out nice last time. But we can probably just, maybe I'll just do that, push it up there and then I know we'll move on.

I do have a quick thought on the Salesforce PR strategy, but we can talk about that async.

Further iterations on this new relationship map — I think I sent that around, you guys saw the potential. We're going forward with agent work, really focusing our efforts on that today. And building for this new approach by the draft to review. And then I think that's really — I'm chatting with Krista from Lima Data about some new stuff. They want to — they're beta-ing some new features from the health side. It's here for good use. I think it appears Slack to DevCon was supposed to give, and then Salesforce. Does it need to make them faster, or is there a way of just trying to improve the efficacy of the system as a whole?

Otherwise, I'll speak to you guys throughout the day. Cam, at 10:30 — both of you, just a review on that PR. I already sent it there. Will talk to you later. Have a good rest of your morning. Thanks, team.

**System Audio:**
We've got to make some good news. All of the indexes — reindexing, default enrichment, analytics jobs — all seem to be working really quickly, including the full reindexing across our companies and people. And even on two workers, it's moving pretty quickly. I'm going to bump it up to four, because I'm also using that kind of dirty set batch system. And I actually just watched it kick off doing a full index too. But just wanted to let you guys know.

That's a great thing. Thank you.

Fun — baby's first React component. I just think it's useless. Telling it not to suggest deleting files and using anonymous connection requests. I'm just like, stop.

I'm hitting my Cloudflare tunnel. Test coverage is still at 7%, so kind of weird because I was moving on the road of deprecations where you can't have multiple files. And so the new code uses the new single file configuration, but the old code still does not, which has kind of worked out nice actually, because that will allow me to segregate. I can use the deprecated tag functions or the Salesforce suggested ones. So I think that was kind of a nice setup for hopefully future use to get rid of the old stuff as we move forward with new stuff.

So hopefully — I don't know the use of those other SmallWorld server instance underscores. I put a comment in that they're not being used anymore, but that would be nice to see. If we're gonna do like a 2.81 dev branch, maybe perhaps that kind of worked out nice last time. It won't be very difficult.

So for me, yesterday I spent some time with David. Had to move it to a relationship map. So I think next day or so, draft to review.

Then I think that's everything. Oh, I'm chatting with — well, I'm making a chat with Krista. More DevCon as opposed to looking into the next kind of sidekick thing, which is Salesforce sync and queues — 160,000 jobs.

And does it need you guys throughout the day? Cam at 10:30, both of you at four, and then — cool. Okay.

The team's morning standup covered solid infrastructure progress — indexing, enrichment, and analytics jobs are all running quickly across companies and people, with plans to scale from two to four workers. Cam is wrapping up Rally (LLM-powered suggested prompts), which David wants to demo on the Postman call this afternoon, and landed a dirty set fix for the indexing job. Todd spent the day fighting Claude's file-deletion behavior and configuration issues but has test coverage at 87%, the Cloudflare tunnel working again, and a path toward deprecating old multi-file config code. The third team member is iterating on a new relationship map (draft to review soon), pivoting toward agent-focused work, and syncing with Krista from Lima Data on new health-side beta features, with Salesforce sync and its 160K-job queue on the horizon.