Audita polishes transcript JSON using glossary-aware and LLM-backed correction modules. It validates proposed corrections and produces corrected text plus reports for review and subprocess integration.

Quickstart

From a checkout, build with Go 1.25.5 or newer:

go build -o ./bin/audita ./cmd/audita

Processing requires a reachable compatible LLM endpoint and access to a model you select. The minimal config targets an unauthenticated local endpoint; edit its endpoint and local profiles to match your service and models. The service example demonstrates a named credential reference and a shared backend. Set its referenced environment variable before validating it. The testing overlay selects both role profiles, but model cost, speed, and quality depend on your provider choices.

Check configuration and profile selection offline:

./bin/audita config validate --config examples/minimal-config.yml
./bin/audita config print-effective --config examples/minimal-config.yml --view profiles

Run the included small input with the minimal config:

./bin/audita process examples/tiny-transcript.json \
  --config examples/minimal-config.yml \
  --glossary examples/tiny-glossary.yaml \
  --output corrected.json

This makes LLM requests. Explicit setting flags override the selected file and optional overlay; general environment setting overrides are no longer read. By default the command writes errors.yml, recoveries.yml, and report.md beside corrected.json. Use --reporting=false to suppress those files, or --debug to retain detailed run artifacts under the configured debug work directory. Model failures affecting some operations can still yield a completed transcript with applied corrections; the summary and reports identify them. See the v2 migration guide for old configurations.

Documentation

Audita is licensed under the GPL v3.0 or later.

Open Source attribution notices are available.