Run the review loop and read status¶

This guide shows how to run the MVP-1 creator/reviewer loop with the fake runner and how to interpret the podcast status output.

1. Run the loop¶

By default, podcast review uses the local Codex/Claude CLIs configured in episode.yaml or ~/.config/podcast-pipeline/config.yaml and writes protocol state under copy/protocol/<asset_id>/.

podcast review \
  --workspace ./workspaces/ep_001 \
  --episode-id ep_001 \
  --asset-id description \
  --max-iterations 3

If you do not have the CLIs installed, use the fake runner instead:

podcast review \
  --fake-runner \
  --workspace ./workspaces/ep_001 \
  --episode-id ep_001 \
  --asset-id description \
  --max-iterations 3

Notes:

--workspace must not exist; the command creates it.
--asset-id must match ^[a-z][a-z0-9_]*$.
Agent CLI configuration is documented in reference/configuration.md.
On convergence, the command prints the selected asset path.
Review iterations are written under copy/reviews/<asset_id>/iteration_XX.<reviewer>.json when the reviewer label is set.
The loop stops only when the reviewer returns ok and the creator reports done, or when max_iterations is reached.

2. Inspect loop status¶

podcast status --workspace ./workspaces/ep_001

Example output:

Workspace: /absolute/path/to/workspaces/ep_001
Asset: description
  Iteration: 2/3
  Verdict: ok
  Outcome: converged
  Blocking issues: none
  Outstanding issues: none

If no protocol state files exist, podcast status reports that no state was found under copy/protocol/.

3. Status fields¶

Asset: the asset id the loop is drafting.
Iteration: current/max; 0/max means no iterations are recorded yet.
Verdict: the latest reviewer verdict (ok, changes_requested, needs_human, or none).
Outcome: in_progress, converged, or needs_human; a decision reason appears as Outcome: ... (reason=...).
Blocking issues: count of issues with severity=error.
Outstanding issues: total issue count plus per-issue lines with severity, message, and optional code or field.