Directing @yagni · Lesson 3 of 4
How do you ask @yagni for help?
Published June 9, 2026
Use the pinned @yagni composer on Front or the contextual rail on most other working surfaces. Command-K or Control-K takes you to the available conversation. One identity, one memory. Its answers are grounded in the same live rows the surfaces read, and in the Library, with citations you can open.
Where do you talk to @yagni?
Two doorways, one voice. Front has a pinned @yagni composer. Most other working surfaces have a contextual rail scoped to what is in the center: open an item and @yagni is ready to discuss that item, open a Brief and it is ready to discuss that snapshot. Command-K or Control-K takes you to the available @yagni conversation; from rail-less Library it carries you to Front.
The scoping is the feature. You never re-explain where you are or paste context in. “Why did this stall?” means the thing on your screen, and the answer arrives already grounded in it.
What makes its answers worth trusting?
The same thing that makes the Front worth trusting: citations. @yagni answers from the live rows of your connected tools and from the Library, the shared store of docs and context the Teams read and write. Every claim traces to a source you can open. When the evidence is thin, it says so and tells you how confident it is, rather than rounding a guess up to a fact.
It is also not built to flatter. Ask “should we ship this week?” and you get its read, including the part that disagrees with you, because an agent that just confirms your priors is a mirror, not a colleague.
What kinds of asks work well?
Three shapes cover most of what operators actually use:
- Read it for me. “Where do we stand with Acme?” “What changed since Friday?” Synthesis questions, answered from everything connected, cited.
- Dig into it. “Why did churn tick up?” “Which deals are blocked on engineering?” Investigation that crosses tools, the kind that used to be an afternoon of tab-hopping.
- Prepare it. “Draft the renewal email.” “Put together context for tomorrow’s call.” The result lands as real work: a draft waiting in the Feed, a doc filed in the Library.
The longer you work with it, the more the third shape dominates, because the Playbook keeps making the drafts arrive closer to done.
What should your first ask be?
Something you actually do not know. Not a test question with a known answer, but the real “what am I missing this week?” Then open the citations and check it. Trust built that way holds, because it was earned against your own data. From there, the rhythm builds itself: read the Brief, clear the Feed, and ask @yagni whenever the next question would otherwise become a meeting. If you want to see it on your own workspace with someone driving, book a call.
Common questions
Is the Front composer a different agent from a contextual rail?
No. There is one voice with one memory. The Front composer and contextual rails are different doorways into the same @yagni conversation model.
What grounds @yagni's answers?
The live rows of your connected tools, and the Library: the shared store of docs and context the Teams read and write. Answers cite their sources, and any citation opens the real thing.
Will it just agree with me?
No. It gives you its read and how confident it is. When the evidence cuts against you, it says so, and when it is not sure, it says that too and hands you the call.
Can @yagni do things, or only answer?
Both. Ask it to draft, prepare, or investigate and the result lands as real work: a draft in the Feed, a doc in the Library. Anything consequential ships only with your approval, as everywhere in YAGNI.