infoPipeline

Run the tests

python3 -m unittest discover -s tests -t .

Run the application

Run the default configured profile once:

python3 planner.py

Run a specific configured profile:

python3 planner.py --profile techno-releases

Validate the default configuration without running the pipeline:

python3 planner.py --validate-config

Validate a specific profile without running the pipeline:

python3 planner.py --validate-config --profile techno-releases

Configuration validation checks that the selected profile can be loaded and assembled. It does not call providers, write a ledger, or send Telegram messages, so it is the safest first check after editing config/pipeline.yaml, prompt paths, templates, profiles, providers, or delivery configuration.

Show the application version:

python3 planner.py --version

--version exits before any pipeline or validation work runs.

Configuration

The default pipeline is assembled from config/pipeline.yaml. Topic selection happens through profiles: each profile supplies the Researcher prompt, Curator prompt, Writer prompt, and Writer template paths. Prompt filenames may be topic-specific. To retarget the pipeline, select a different profile or add one that points at different prompt files rather than changing Python source code.

Source-backed Researcher providers can also declare stage-level discovery criteria. The checked-in Bandcamp Researcher configuration uses discovery to describe its Bandcamp Discover collection criteria; omitting that block keeps the provider's default discovery behavior. Bandcamp discovery may also be an empty object or a partial object, such as tag_norm_names plus size; omitted documented fields are filled from Bandcamp provider defaults before collection.

More than one Researcher can run before Curator by repeating flat name: researcher entries in the top-level stages list. Each Researcher uses its own provider configuration. The normalized items from successful repeated Researchers are combined in stage order, duplicate URLs keep the first occurrence, and Curator receives one candidate pool. Do not nest Researcher instances under fields such as researcher.instances, researcher.providers, or discovery_sources.