‹ Product recipesProduct · Agent Blueprint
Feature launch brief
Everything that was promised, debated, and decided about a feature before it ships, assembled with the customer evidence behind it.
Why this is memory-only: a complete launch brief requires the customer promises, debates, and rejected tradeoffs that do not survive in ordinary artifacts.
Side by side: token usage, with and without Paperbase
Memory onlyWithout Paperbase
Not recoverable from artifacts alone
- Collect the spec and the launch plan.
- Ask sales what was promised to which customers.
- Reconstruct the design debate from whoever was in the room.
- Gather the roadmap items the feature displaced.
- Ship without a complete picture of what was committed.
With Paperbase
Persistent memory layer
- Pull the feature's Decision history, with rationale.
- Pull Commitments made about the feature - to customers, internal, and to the roadmap.
- Pull the customer Accounts and quotes behind the feature.
- Assemble the launch brief with everything cited.
Agent prompt
You are assembling a launch brief for {{feature_name}}.
Using Paperbase:
1. Pull every Decision about {{feature_name}}, in order, with rationale and owners.
2. Pull every Commitment made about it: to customers, to other teams, to the roadmap. Include delivery status.
3. Pull the customer Accounts that requested or reference the feature, with their quotes.
4. Note the roadmap items it displaced or delayed.
Output:
- What we decided and why (chronological).
- What we promised, to whom, and the delivery status of each.
- The customer evidence behind the launch.
- Risks: anything promised that will not be in this release.Placeholders in {double_braces} are inputs the agent will ask for at runtime. Give the prompt to any agent connected to Paperbase, and the rest grounds in your own memory graph.