Follow-up work for slash commands and targeted messages

[nickwalkmsft]

• Everything about prompt preview needs to go back in when ready, including guidance.
• Needs info and guidance about reply() and related methods, including how they interact with prompt preview, and best practices about using them to control response visibility
• • Should build out a dedicated how-to page for the private-and-make-public workflow that brings together targeted messaging, prompt preview, and suggested actions, and link it from these.
• Needs additional screenshots and a scrub to ensure they evoke specifically what the surrounding content is talking about.
• Slash commands article doesn't discuss "triggers": ["mention"] at all. See Full reconciliation of prompt suggestions with commanding: overall docs organization, storytelling, other items #14380
• Need better conceptual explanation of how sending works - why does WithRecipient even allow you to set isTargeted: false and what does that do?
• Details on exact nature of TM support for reactions, replies and forwarding (see "Key aspects" section)
• Maybe need REST samples
• "Update or delete a targeted message" and maybe a couple other spots ultimately shouldn't have snippets, it should direct people to a Teams SDK-based messaging guide that shows those standard operations, but our overall messaging guide isn't in Teams SDK, it's in Bot Framework. This is a high priority target when moving the Teams SDK docs over to Learn.
• Slash commands article could use a few concrete scenarios sprinkled in. I removed the examples of an agent providing commands named "/help" or "/settings" because those are kind of confusing in the context of built-in commands, we should give scenarios with something more interesting.
• Include explicit technical instruction in addition to "see following example" - it's bad practice and particularly bad for coding agents to only provide an example that looks like the right result, without explicitly saying "use such and such property".
• Need blanket guidance for slash commands, including opinionated platform-POV guidance about leaning in to agents' ability to handle natural language requests and not only providing commands.
• Removed code sample section from slash commands page, it looks like we don't have one yet.
• Consider changing to use zone pivots instead of the tab boxes

See #14329 for related work in flight.

Related: #14380

[microsoft-github-policy-service]

Hi @nickwalkmsft! We will investigate and if we require further information we will reach out to you. Please use this link to escalate if you don't get replies in two working days.

Best regards, Teams Platform

[m365-agent-toolkit-assistant]

👋 Hi @nickwalkmsft,

Thank you for reporting this issue! We want to let you know that:

⏱️ Our team will personally review your issue
🎯 You can expect a response within 1 business day

✨ This is an AI-generated initial response
🤖 AI-Generated Preliminary Analysis:

Classification: Feature request — this issue is asking for doc expansion/reorganization and additional guidance, not reporting broken product behavior.

• The body is a documentation backlog: “needs info and guidance,” “should build out a dedicated how-to page,” “maybe need REST samples,” and “consider changing to use zone pivots.”
• Learn already states that manifest commands are used for prompt starters, and that agents must use type: prompt with a valid prompt value.
• Learn also already documents targeted messages as preview, including restrictions on reactions/replies/forwarding and SDK usage of WithRecipient(..., isTargeted: true) / withRecipient(..., true).

Existing coverage:

• Closely related open docs issues already exist and appear to be the main tracking work.

What to do now:

1. Treat this as docs-planning work, not an engineering bug.
2. Consolidate under the broader command/prompt reconciliation issue if you want one parent tracker.
3. Add explicit cross-links between prompt starters, slash commands, and targeted messages pages.

Useful links:

• MicrosoftDocs/msteams-docs#14380 — Full reconciliation of prompt suggestions with commanding overall docs organization, storytelling, other items
• Create prompt suggestions
• Enable targeted messages for agents

If helpful, I can turn this into a tighter doc action plan.

---

Note: This is an automated first response generated by AI. A human team member will review your issue and provide a more detailed response soon. We appreciate your patience!

[nickwalkmsft]

FYI @dampbrownshoes