Governance metrics dashboard — coverage by category, cross-reference density, compliance readiness, and project comparison across all working projects
Scanned 6/11/2026
Install via CLI
openskills install tractorjuice/arckit-codex---
name: arckit-graph-report
description: "Governance metrics dashboard — coverage by category, cross-reference density, compliance readiness, and project comparison across all working projects"
---
# Graph Report — governance metrics across all projects
You are producing a **governance metrics dashboard** for all working projects in the repo. The hook has already done all the computation — your job is to render it cleanly.
**This is a read-only diagnostic command. No files are written.**
## User Input
```text
$ARGUMENTS
```
(No arguments — graph-report always scans every working project under `projects/`, excluding `000-global`.)
---
## What This Command Does
The hook scans every working project and computes:
| Section | Meaning |
|---|---|
| **Project Comparison** | One row per project: artifact count, cross-reference count, density, compliance readiness |
| **Coverage by Category** | Per-project matrix of how many doc types are present in each category (Discovery, Planning, Architecture, Governance, Compliance, …) |
| **Compliance Readiness** | Per-project list of which HIGH-severity doc types are present and which are missing |
| **Cross-Reference Density** | Edges per artifact — a proxy for traceability strength |
The hook delivers this as a structured pre-processor block (`## Graph Report Pre-processor Complete (hook)`).
## Process
### Step 1: Read the hook output
The hook block at the top of your context contains:
- A summary line (projects scanned, total artifacts, total cross-references)
- A project comparison table
- A coverage-by-category matrix
- A per-project compliance readiness breakdown
- A density interpretation legend
If the hook block is missing (unlikely — it always fires for `$arckit-graph-report`), inform the user and stop.
### Step 2: Render the dashboard
Output to the console (no file written) using this structure:
```text
📊 ArcKit Graph Report — <N projects, M artifacts, K cross-refs>
PROJECT COMPARISON
<render the project comparison table from the hook>
COVERAGE BY CATEGORY
<render the category matrix>
COMPLIANCE READINESS (HIGH-severity doc types)
<render the per-project breakdown>
DENSITY INTERPRETATION
<render the legend>
🔎 Outliers
- Lowest compliance readiness: <project, percentage>
- Lowest cross-reference density: <project, density>
- Projects missing whole categories: <list, if any>
▶ Recommended actions
- <bullet list — for each project with <50% compliance readiness, suggest the top missing $arckit-* command>
```
### Step 3: Stop
No file is written. The output is the deliverable.
---
## Notes
- **Compliance readiness** counts the presence of these 12 HIGH-severity doc types: TCOP, SECD, SECD-MOD, DPIA, SVCASS, RISK, TRAC, CONF, PRIN-COMP, AIPB, ATRS, JSP936. Not all are required for every project — UK Gov projects need more of them than internal-only ones — but the dashboard surfaces gaps so the architect can decide.
- **Density** is `edges / nodes` per project. Higher density means better-connected artifacts. The legend in the hook output gives interpretation guidance.
- For a single-project repo the comparison table has one row — the report is still useful for category coverage and compliance readiness in that case.
- `000-global` (architecture principles) is excluded from the per-project rows but principles still inform the artifacts via cross-references in working projects.
## Related commands
- `$arckit-navigator <project>` — single-project "what's next" GPS, complementary to graph-report.
- `$arckit-analyze <project>` — deep governance analysis for one project (sections, requirements, principles, risks).
- `$arckit-health` — point-in-time health rules across all projects.
No comments yet. Be the first to comment!