Skip to Content

rag-forge report

Generate a pipeline health report dashboard

Synopsis

rag-forge report [options]

Description

report aggregates existing pipeline data — indexed chunk counts, the most recent audit JSON report, and drift baseline status — into a single HTML health dashboard. The generated file gives you a quick overview of your pipeline’s operational state without requiring any new LLM calls.

Unlike audit (which runs evaluation and scores your pipeline) and assess (which checks maturity criteria), report is purely a presentation layer over data that has already been collected. Run it after an audit to produce a shareable artefact, or on a schedule to keep a persistent health record.

The command delegates to the Python rag_forge_evaluator.cli module. The output HTML file is written to the directory specified by --output.

Options

FlagDefaultDescription
-o, --output <dir>./reportsOutput directory for the generated report
--collection <name>rag-forgeName of the vector store collection to inspect for chunk counts

Examples

Basic usage

rag-forge report

Write to a custom output directory

rag-forge report --output ./docs/health

Report for a named collection

rag-forge report --collection my-project --output ./reports/my-project