// dev_with_ai --week 2026-09-07
Remote coding tools, model reactions and five public API changes: inline filter results, broadcast stats, external purchases, subscription dates, price errors.
// code_support
Building something on ClickFunnels? Share your use case with the Funnel Coders Discord community and get support from ClickFunnels engineers.
Share your use case// from_the_wild
What the team read, watched and argued about this week.Run tests remotely without committing unfinished work first, and get the remote exit status back in your shell. The part people were curious about was making remote work feel transparent. - Engineering Team
docs.humanlayer.com/guide/automation-sessions ↗
Run a coding session from CI or a scheduled job, with progress visible in HumanLayer Cloud. It came up as another approach to the remote-compute problem people were already experimenting with. - Engineering Team
See local and remote Herdr sessions together, with agents needing attention first. The reaction was that these tools are approaching the same coordination problem in different ways. - Engineering Team
Use existing model subscriptions to run coding agents from one open-source app. It joined the comparison of different ways to move agent work between machines. - Engineering Team
x.com/mitchellh/status/2093451043661316217 ↗
The macOS preview uses a custom binary protocol with terminal state replicas on the server. The demo's fast startup got an enthusiastic response. - Engineering Team
anthropic.com/claude-fable-and-mythos-5-1 ↗
Anthropic says Low and Medium can match or beat Fable 5 at lower cost, so try them before settling on High. Early impressions included better efficiency and a more natural tone, though the budget stories were mixed. - Engineering Team
Anthropic suggests asking, “Please remove all mannered prose.” A couple of us already liked how 5.1 was talking, and this tip got an approving response. - Engineering Team
x.com/skirano/status/2095595932335170031 ↗
The early-access demos turn an image into an animated 3D model and a video into interface code. We saw excitement about the results, plus some skepticism about launch hype and token costs. - Engineering Team
// shipped
New on the public API, with changelog or API documentation links.developers.myclickfunnels.com/changelog ↗
Every refine-filter endpoint takes a results parameter and returns the matching ids, or a projection of the fields you name, inline. One call instead of a filter plus a fetch. - Engineering Team
developers.myclickfunnels.com/changelog ↗
Delivery, engagement and revenue numbers for a broadcast at GET /api/v2/emails/broadcasts/{id}/stats, and the broadcast itself now exposes its split-test variants and the winner. - Engineering Team
developers.myclickfunnels.com/changelog ↗
Set up a listener for ClickBank or CopeCart purchases without opening the app: list and create the incoming endpoints, then map external plans to your prices. - Engineering Team
developers.myclickfunnels.com/changelog ↗
PATCH /api/v2/orders/:id with trial_end_at or renewal_at takes an absolute timestamp, so a support script can extend a trial instead of a human clicking through the order. - Engineering Team
developers.myclickfunnels.com/changelog ↗
Invalid price payloads now come back as a 422 naming the field, where they used to be a generic 500 that looked the same whatever was wrong. - Engineering Team
$ build_with_us --submit
Webhooks, API integrations, funnel automation: share what you are building with the Funnel Coders Discord community and get support from ClickFunnels engineers.
Share your use case// dev_with_ai --week 2026-09-07 · all posts