code-partition-stats

name	code-partition-stats
description	Generate per-partition code statistics and before/after diffs for refactor impact analysis. Use when evaluating architecture changes (for example web vs desktop vs core consolidation), creating baseline metrics before a large refactor, or reporting quantitative progress using file and line deltas.

Code Partition Stats

Generate reproducible measurements for codebase partitions and compare refactor impact over time.

Quick Start

From repository root, create a baseline snapshot:

python ".opencode/skills/code-partition-stats/scripts/partition_stats.py" snapshot \
  --root "." \
  --output ".sisyphus/evidence/refactor-baseline.partition-stats.json"

After changes, create a second snapshot:

python ".opencode/skills/code-partition-stats/scripts/partition_stats.py" snapshot \
  --root "." \
  --output ".sisyphus/evidence/refactor-after.partition-stats.json"

Build a quantitative diff report:

python ".opencode/skills/code-partition-stats/scripts/partition_stats.py" diff \
  --before ".sisyphus/evidence/refactor-baseline.partition-stats.json" \
  --after ".sisyphus/evidence/refactor-after.partition-stats.json" \
  --output ".sisyphus/evidence/refactor-diff.partition-stats.json"

Default Partition Model (mps-wall-refactor)

The bundled script includes this preset:

• web_frontend: apps/web/src
• desktop_frontend: apps/desktop/src
• core_shared: packages/core/src
• wall_feature:
  • apps/web/src/wall
  • apps/web/src/scene/wall-scene.ts
  • apps/web/src/scene/route.ts
  • apps/desktop/src/wall
  • apps/desktop/src/features/platform/wall-platform-adapter.ts
  • packages/core/src/wall

Custom Partition Config

Pass --config <json-file> to override partitions/extensions/excluded directories.

Example config file:

{
  "partitions": {
    "web_frontend": ["apps/web/src"],
    "desktop_frontend": ["apps/desktop/src"],
    "core_shared": ["packages/core/src"],
    "wall_feature": [
      "apps/web/src/wall",
      "apps/desktop/src/wall",
      "packages/core/src/wall"
    ]
  },
  "extensions": [".ts", ".tsx", ".js", ".mjs", ".css"],
  "exclude_dirs": ["node_modules", "dist", "dist-release", "release", ".git", ".sisyphus"]
}

Run with custom config:

python ".opencode/skills/code-partition-stats/scripts/partition_stats.py" snapshot \
  --root "." \
  --config ".sisyphus/evidence/partition-config.json" \
  --output ".sisyphus/evidence/refactor-custom.partition-stats.json"

How To Read Results

For refactor outcomes, prioritize:

1. non_empty_lines delta by partition (primary size signal)
2. file_count delta by partition (surface-area movement)
3. wall_feature movement from app partitions to core_shared
4. extension-level deltas to detect accidental scope expansion

Use the diff JSON as a quantitative appendix in refactor PRs.