Update instructions for steering agent conversations#43829
Update instructions for steering agent conversations#43829Sharra-writes merged 5 commits intogithub:mainfrom
Conversation
Clarified instructions on interacting with the agent while it is processing a task, emphasizing the absence of a separate instruction queue and how to steer the conversation. The origianl was vauage as to how to actually steer it, and new to the CLI users might not understand. Also, queuing was not explained as it was assumed to be naturally understood. Not so. This should smoothing the instructions over without undue weight of words.
|
Thanks for opening this pull request! A GitHub docs team member should be by to give feedback soon. In the meantime, please check out the contributing guidelines. |
How to review these changes 👓Thank you for your contribution. To review these changes, choose one of the following options: A Hubber will need to deploy your changes internally to review. Table of review linksNote: Please update the URL for your staging server or codespace. The table shows the files in the
Key: fpt: Free, Pro, Team; ghec: GitHub Enterprise Cloud; ghes: GitHub Enterprise Server 🤖 This comment is automatically generated. |
github-actions
bot
added
the
triage
There was a problem hiding this comment.
Pull request overview
Updates the “Steer agents” docs to better explain how to steer an in-progress {% data variables.product.prodname_copilot_short %} CLI agent task, clarifying how additional input is handled while the agent is working.
Changes:
- Replaces the prior “follow-up / instruction queue” phrasing with more explicit guidance about entering prompts while the agent is running.
- Explicitly states there is no separate instruction queue and explains message ordering during an active task.
…er-agents.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
…er-agents.md Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This comment was marked as spam.
This comment was marked as spam.
Co-authored-by: Sharra-writes <sharra-writes@github.com>
|
@Brenduns I think this is clearer than what we have, but not so different or consequential that I need to get SME approval. Which is my favorite kind of update, to be honest. I cleaned up a couple of adverbs just to streamline things a bit, but otherwise it looks good. I'll just get through these checks and hopefully get it merged. |
|
Thanks very much for contributing! Your pull request has been merged 🎉 You should see your changes appear on the site in approximately 24 hours. If you're looking for your next contribution, check out our help wanted issues ⚡ |
Clarified instructions on interacting with the agent while it is processing a task, emphasizing the absence of a separate instruction queue and how to steer the conversation.
The original was vague as to how to actually steer it, and new to the CLI users might not understand. Also, queuing was not explained as it was assumed to be naturally understood.
Why:
Ad-hoc find of vague instructions, and fix to improve them so new to CLI users will understand how to interact.
Closes:
What's being changed (if available, include any code snippets, screenshots, or gifs):
Clarified instructions on interacting with the agent while it is processing a task via the same prompt and then emphasizing the absence of a separate instruction queue. Original presentation hinted at some area to queue other instructions and omitted the basic detail that additional instructions are processed when entered as part of the ongoing task. That is only obvious when you already know it.
Check off the following: