# Delivery Governance

> Use for the state contract of a delivery run — the `.kaizen/runs/<run-id>/run.yaml` six-field state file and its append-only `journal.md`: reading a run's current stage/step/role/result before doing delivery work, recording the result of an action or checkpoint (pass, pass_with_risks, blocked, failure) and deciding where it routes next, or checking that `run.yaml` and its evidence were updated together. Applies at every run's target resolution and every transition, not one specific stage.

- Skill: `bhangeef16/delivery-governance` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add bhangeef16/delivery-governance`
- Raw SKILL.md: https://api.skillmd.com/api/skills/bhangeef16/delivery-governance/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: BhangeeF16 (https://skillmd.com/u/bhangeef16)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/bhangeef16/delivery-governance

---


# Delivery Governance

This is a meta-skill: a rule that applies across every delivery run rather than belonging to one action. Use it whenever a run's state is read from or written to — its position, result, and blocking questions. The content below is read live from this plugin's own `skills/delivery-governance/delivery-governance.md` every time this skill runs, so there is no separate paraphrased copy anywhere for it to drift against.

## Delivery governance contract (live)

!`cat "${CLAUDE_PLUGIN_ROOT}/skills/delivery-governance/delivery-governance.md"`

## Instructions

1. Before doing any delivery work on a run, read its current state from `.kaizen/runs/<run-id>/run.yaml` exactly as the State Contract above defines the format — the six fields in the stated order, `blocking_questions` as a YAML list.
2. Advance state only through `kaizen.py run advance` (`bin/kaizen_run.py`), in the same change set as the evidence it describes — it writes `run.yaml` and appends the matching `journal.md` line together; a hand-edited `run.yaml` or a state record that outruns (or lags) its evidence is a defect.
3. At scored checkpoints, apply the `evidence-verification` skill per the Evidence Verification section above — do not reintroduce a blind re-scoring dispatch alongside it.
4. Resuming a run after any gap or `/compact`: re-read `run.yaml` and `journal.md` under that run's own `.kaizen/runs/<run-id>/` before dispatching or re-dispatching anything.
5. If you are re-reading this skill after `skills/delivery-governance/delivery-governance.md` changed, trust what you just read above over any memory of a previous run — that live content is always the current source of truth for how delivery state is read and written in this workspace.