coding_charge_reviewer independently reviews a coding agent’s proposed CPT/ICD/modifier output against the clinical documentation and flags errors. It is part of the Nanonets Coding Engine tool group.
The strict reviewer system prompt is baked into the tool.
Inputs
documentation(required): the clinical documentation the codes were derived from.proposed_codes(required): the generated charges and coding suggestions from the CPT step to review.diagnosis_output(optional): the diagnosis-step ICD-10 output, for extra context.
Output
agreed=true only when the reviewer would sign off on submitting the claim.
Human-readable text
In the task feed the tool renders a## Coding Charge Reviewer table (agreed, issue count) and an ## Issues table (severity, code, problem, suggested fix). The full result struct stays in structuredContent.
Agent Instruction Guidance
Mergeissues into the final validation_issues array with source verifier.
Notes
- LLM tool. Billed per call.
default_enabledis false.