Project thread
Feature: new mobile onboarding flow. Sarah has asked for input from the design lead and the iOS team before Friday.
Use Case
RenX for project collaborations
Keep outside input attached to the project it belongs to — with context-aware drafting, explicit approval, and synthesis back into one thread.
What you just watched
The flow starts inside a project conversation — not in a blank chat box. That matters. Your agent has the spec, the history, and the context already attached. When it drafts an outreach to a collaborator, the question arrives fully framed.
You approve exactly what leaves the project. The reply comes back, from whatever channel the collaborator prefers, and lands back inside the project thread. If there are multiple collaborators, the agent synthesizes their replies into a decision-ready summary with each original source linked inline — so the synthesis never becomes a game of telephone.
This is the pattern that breaks when teams fragment across tools. A design review in Figma comments, an engineering reply in Slack, a legal note in email. Decisions get made, but the trail disappears. RenX keeps the project as the system of record.
Project context Approval gates Multi-channel replies Inline synthesis
A concrete scenario
Sarah is a product manager driving a mobile onboarding redesign. The new flow touches four teams: design for the visual language, iOS engineering for the platform-specific permission handling, marketing for the in-app copy, and analytics for the funnel measurement. The launch is Friday.
Without a coordination layer, Sarah writes four separate messages in four different places. Design replies in a Figma comment. iOS replies on Slack. Marketing replies in a shared doc. Analytics replies by email. By Thursday afternoon, the decisions are made but nobody can quickly see where clause X came from — and when the launch is retrospectively reviewed, the trail is scattered.
With RenX, Sarah's agent drafts the outreach to each collaborator with the right slice of context for their review, Sarah approves each draft, and every reply lands back in the project thread. When the launch is reviewed a week later, the entire decision path is one thread — not four tools and six searches.
Who this flow is for
Product managers Running cross-functional review without losing the spec, the rationale, or the replies that shaped the decision.
Founders Pulling design, engineering, legal, and analytics into one project conversation before a launch or major change.
Operators Keeping outside input attached to the work itself instead of scattering it across chat, email, and shared docs.
Same flow as a real thread
Here is the onboarding scenario rendered as an actual project conversation. Each outbound message required approval; each reply came back from a different channel; the synthesis keeps both originals linked.
A mobile onboarding redesign running cross-functional review. Drafted by the agent, approved by the PM, replies from Slack and Telegram, synthesized back into the project thread.
Your agent
Pulled the spec and recent thread history. Drafting outreach to Jules (design lead) with the core question and the annotated mockup. Approve when ready.
You
Approve. Also send the iOS-specific flow to Min.
Jules · Slack
Onboarding step 3 should come before the permission prompt. Happy to mock a v2 — send me the analytics.
Min · Telegram
iOS 17+ can do the notification permission inline. Older versions need the pre-prompt. Split the flow by OS.
Your agent
Two replies in. Synthesis: reorder step 3 before permission prompt (Jules), split iOS 17+ vs older (Min). Both threads linked in the project. Ready to draft the v2 spec?
How the flow runs
- Your agent reads the project context first. Before drafting anything, the agent pulls the relevant thread history, attached docs, and specs tied to the project so outside input is framed with the right background — not raw questions stripped of their setting.
- You approve what leaves the project. The agent shows you the outreach draft and exactly which context will be shared with the collaborator. You edit, approve, or reject — nothing leaves without your sign-off.
- External replies return to the project thread. Whether the collaborator responds on Slack, Telegram, WhatsApp, or email, the reply lands back inside the same project conversation — not in a separate inbox that someone has to remember to sync.
- The agent synthesizes feedback into a decision-ready summary. Multiple collaborators, multiple replies, one synthesis. The agent keeps the original threads linked so you can always trace a claim back to its source.
Platform fit
- Project workspaces keep the spec, history, and attached files in one scoped conversation
- Channel support so each collaborator can reply where they already work
- Approval gates so you control exactly what context leaves the project
- Synthesis with sources so summaries never lose their link to the original reply
Configure the pieces
FAQ
Does the agent share the whole project with external collaborators?
No. It only shares the specific context you approve for that outreach — typically the question, a short setup, and any attachment you explicitly include. The full project history stays internal.
How does the synthesis work without distorting what each collaborator said?
The agent produces a summary but keeps each original reply linked inline so you can verify any point. The synthesis is framed as "here is what was said", not as "here is what I think they meant".
Can I route different questions to different collaborators in the same project?
Yes. The project thread can hold multiple parallel collaboration sub-threads — one for engineering review, one for design, one for legal — all visible in the same project view.
Does this work for cross-functional teams?
Yes. This is the use case it was built for — product managers, founders, and operators who need input from people outside the current working group without fragmenting the decision trail.
Keep outside input inside the project.
Configure your workspaces, approvals, and channels first, then let your agent run the cross-functional review without fragmenting the decision trail.