What it is
Work Request Parser is the front door of the work homebase: it converts a hunch into a task the system can actually execute.
Homebase Building Block
Work Request Parser.
A translation layer that turns vague human requests into executable work orders with goals, data needs, constraints, missing context, permissions, and output format.
Messy request -> executable work order
The system turns a vague ask into work it can run.
Calls server parsing with rate limits.
Goal
Identify which providers should be surfaced differently.
Audience
Clients who cancel
Data needed
Cancellation, scheduling, and provider-performance data.
Success metric
Metric unclear.
Permission needed
Read scheduling notes before proposing changes.
Output format
Short plan, missing context, and first recommended action.
Missing context
- No timeframe specified.
- Metric unclear.
- Need source of cancellation data.
- Need permission to inspect scheduling notes.
The first product skill is turning intent into a task the system can actually execute.
Problem
Mainstream users ask for outcomes in messy language. Agents need a clearer task shape before they can gather context, ask for permission, or produce useful work.
How it works
- Parse the raw request into goal, audience, data needed, constraints, success metric, missing context, first step, permissions, and output format.
- Show parser uncertainty before execution begins.
- Turn missing fields into questions the user can answer or approve.
Why it matters
- Clear work orders reduce wasted agent motion.
- Missing-context warnings make vague intent easier to repair.
- The parser preserves user language while creating enough structure for execution.
Behavior
Good behavior
The system converts a cancellation question into goal, audience, data needed, metric, missing context, permission need, and first step.
Bad behavior
The system treats a vague ask as complete and starts working in the wrong direction.
Recruiter signal
This shows product judgment around trust, AI UX, systems thinking, and the difference between useful automation and opaque automation.