openlogos feature

Features group a module’s scenarios into user-facing capabilities (F01, F02, …). They are stored in logos/logos-project.yaml — features[] entries {id, name, module, spec?} plus a feature key on each scenario — and maintained by AI; the CLI only reads them.

Two commands work with features:

• openlogos feature list — a read-only grouped view
• openlogos feature-backfill — writes an AI prompt for grouping scenarios that are not yet in a feature

Synopsis

openlogos feature list [--module <id>] [--format json]
openlogos feature-backfill [--module <id>] [--format json]

Must be run from the project root.

Options

Option	Description
--module <id>	Limit to one module. An unregistered module fails with MODULE_NOT_FOUND.
--format json	Print a JSON envelope

openlogos feature list

For each module, lists every registered feature with its scenarios (and its feature-spec document, if linked). Scenarios that belong to no registered feature appear in a final __ungrouped__ group. A module with neither features nor scenarios shows an empty list.

openlogos feature list --module core

🗂  Feature Groups

  core  Core
    F01  Project setup  → core-01  (2)
      - S01  CLI init
      - S02  Status
    __ungrouped__  …  (1)
      - S07  Watch

This command never writes anything.

openlogos feature-backfill

Generates logos/feature-backfill-prompt.md, an instruction for your AI assistant to cluster ungrouped scenarios — and, in adopted (brownfield) projects, reverse-engineered scenario candidates — into features, allocate IDs, and write the result back to logos/logos-project.yaml.

The command itself does not modify logos-project.yaml. Running it again is safe: already registered scenarios are not offered as new candidates.

  ✓ Generated: logos/feature-backfill-prompt.md
    scenarios=12  ungrouped=4  features=3  baseline_candidates=0

Next step:
  Tell your AI assistant: "Read logos/feature-backfill-prompt.md and execute the instructions"

JSON output carries prompt_path, scenarios_total, ungrouped_total, features_existing and baseline_candidates_total.

Errors

Error	Cause	Fix
PROJECT_NOT_INITIALIZED / logos/logos.config.json not found	Not in a project root	cd to the project root
MODULE_NOT_FOUND	--module is not registered	Run openlogos module list
BASELINE_COMMIT_IN_PROGRESS	A baseline seed commit is still running (feature-backfill)	Retry later
BASELINE_PROVENANCE_INVALID	A reverse-baseline provenance section cannot be parsed (feature-backfill); the error lists the files	Fix the named files, then rerun

Related commands

• module — Manage the modules features belong to
• status — Its JSON output includes feature groups per module when present
• index — Rebuild the resource index in logos-project.yaml