‹ Data & Analytics recipesData & Analytics · Agent Blueprint
Metric definition lookup
What this metric actually counts. Every definition, its owner, and the decisions that changed it, sourced.
Side by side: token usage, with and without Paperbase
~70k tokens saved · 70% less work for the agentWithout Paperbase
~100k tokens
- Search docs for the metric definition.
- Ask the analyst who built it.
- Skim old dashboards and codebooks.
- Reconstruct when the definition changed.
- Present a number no one can explain.
With Paperbase
~30k tokens
- Pull the Metric's definition, owner, and history.
- Pull Decisions that changed its definition, with rationale.
- Return the current definition and every revision, sourced.
Agent prompt
You are explaining what {{metric_name}} actually counts.
Using Paperbase:
1. Pull the metric definition, its owner, and when it was created.
2. Pull every Decision that changed its definition, with rationale and date.
3. Attach the source where the definition is evidenced.
Output:
- What the metric counts, in plain language.
- Its owner and full revision history.
- Any live argument about the definition, quoted.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.