v1.1

Aperture Bot now comes back to you: every finished task ends with what it would do next, and it can look your repository over on its own schedule. Workflows on @v1 get this release too.
New
Suggestions after every task. A finished run, clear or red, ends with up to three things the bot would do next, in its reply under Next, I would suggest. Nothing is done until you ask for one with /aperture, or send it from the Bot page.
Look things over (scheduled: look-over). On the workflow's schedule the bot reads CI on the default branch and the open issues and pull requests, then posts a few sentences on how things stand, with up to three things it would do next, on its own issue ("Aperture Bot: looking things over"). It checks nothing out and changes nothing. Choose it on the Bot page under Setup → Standing jobs → Look over (every night or every Monday), or set scheduled: look-over yourself.
Better
A busy model is retried. When the provider is rate-limited or overloaded (Gemini's free tier often is), the bot waits and tries again, three times over about a minute and a half, before giving up.
Gemini works with new keys. The bot asks for Gemini 3.8 Flash, since Google turns new keys away from Gemini 2.5 Flash. With an Anthropic key it uses Claude Opus 5.5. When a provider refuses a request, the error says what the provider said.
Clearer errors. A model address that answers with something other than a model's reply (a parked address saying "OK", an error page) says what came back and to check the address and model name. A missing model key says where to add it.
Fixed
/aperture check counts whenever "check" is on the first line, so a signature under it no longer turns it into a task.
When the model plans without saying anything, the reply no longer quotes the editor's "Click Build it" as what the agent said.
Built from Hankaws/aperture@ac8b236. Inputs are unchanged apart from the new scheduled: look-over value, so nothing in an existing workflow needs to change.