Repository navigation
feat(api): the first playbook asks whether it runs on its own, and builds the workflow in - #332
Merged
rami-hatoum merged 1 commit intoOct 26, 2026
Conversation
…ilds the workflow in A person asked for a playbook that continuously reviews raw material, and the agent that followed the first-playbook recipe stopped after one run of one reasoning function: the recipe ended at that run and named workflows only in its list of what comes next. Where the server carries the workflow guide, the recipe now says that a first playbook holds the workflow that runs its function where the work should run on its own. After what the playbook is for, the agent learns how its work starts, and takes the person's words as the answer where they already say so. Once the function has answered, it writes that workflow with its trigger as the workflow guide says: cron or every for set times, and on for what arrives, with the material the person pastes published as events of that type until the function's tools reach the sources. It shows the document, saves it once the person agrees, says when it next runs or what starts it, and reads the run an event started with list_runs and get_run. There alone its calls name publish_event, list_runs and get_run. Where the work starts only when someone asks, the recipe ends at the run as before. What comes next is a closing paragraph, after the steps, that names the tools, memory and, where carried, interaction, and first, where workflows are served, the schedule recipe and the workflow guide for changing when it runs. The workflow words extend the insertion the interaction guide already used: each guide adds its insertions and its calls to a recipe only where the server carries it. Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>
rami-hatoum
deleted the
the-first-playbook-asks-whether-it-runs-on-its-own
branch
October 26, 2026 17:15
rami-hatoum
pushed a commit
that referenced
this pull request
Oct 26, 2026
🤖 I have created a release *beep* *boop* --- ## [0.23.0](v0.22.1...v0.23.0) (2026-10-26) ### Features * **api:** the first playbook asks whether it runs on its own, and builds the workflow in ([#332](#332)) ([359baba](ht tps://github.com/BeOnAuto/runtime/commit/359babaeb15e51bf7e38445896318 137b87477c9)) --- This PR was generated with [Release Please](https://github.com/googleapis/release-please). See [documentation](https://github.com/googleapis/release-please#release-please).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
A person who asks their agent for a playbook that keeps reviewing material got a playbook with one reasoning function, run once, that they had to run by hand afterwards: the first-playbook recipe ended at that run and named workflows only among the things to do next, and the schedule recipe, the one place that says a workflow starts on a schedule or on an event, was never opened.
The recipe now says that a first playbook holds a first function the person sees answer and, where the work should run on its own, the workflow that runs it. After the agent learns what the playbook is for, it learns how the work starts: on its own at set times, on its own whenever something arrives, or only when someone asks, and where the person's words already say so ("every morning", "continuously", "whenever a lead comes in") it takes that as the answer. The function is still written, saved and run once first. Then, where the work starts on its own, the agent reads the workflow guide and writes the workflow with its trigger:
cronoreveryfor set times,onfor what arrives, which lets the person start runs today by publishing the material they paste as events and move to a clock once the function's tools reach the sources. It shows the document, saves it when the person agrees, says when it next runs or what starts it, and reads the run an event started. Where the work starts only when someone asks, the recipe ends at the run as before, and what comes next is a closing paragraph that names the schedule recipe and the workflow guide for changing the trigger later.The workflow words are inserted only where the server carries the workflow guide, through the same mechanism that already inserted the interaction guide's words: each guide adds its insertions and the tools they call to a recipe only where it is carried, so a server without workflows serves the recipe unchanged and a prompt is offered only where every tool it calls is listed. The recipe's size stays within the served budget, and the reference page and the first-workflow tutorial describe the new steps.