[wan9chi]

Motivation

Since #966, vp run --dry-run --json fails for a task whose command runs vp run …. The dry-run RFC shows such a step as expandedFrom with its own plan, so the run plan is a tree, and each nested plan has its own dependencies.

Changes

// "nested": "vp run build"
{ "expandedFrom": "vp run build",
  "plan": { "build": { "package": ".", "name": "build", "steps": [ { "program": "…", "cache": { "status": "hit", … } } ] } } }

• The layout recurses into nested graphs. Each plan's dependsOn is scoped to its own tasks, so the same task ID can appear at several levels.
• expandedFrom is the item's command as written, with any args given after the task name.
• A leaf's step is found by its path through the nested plans.
• The error for tasks that run vp run goes away.

A nested vp run --dry-run is still a spawned step, which a dry run doesn't start. The e2e fixture run_plan_json_nested covers plans at several levels, args forwarded to a nested run, and a nested dry run.

🤖 Generated with Claude Code

---

Note

Medium Risk
Touches dry-run JSON output shape and reporter layout for nested graphs; behavior is covered by new e2e snapshots but consumers parsing the old flat step model need to handle NestedStep.

Overview
vp run --dry-run --json no longer errors when a task runs another vp run. Those steps are emitted as nested objects with expandedFrom (the command as written, including args after the task name) and an embedded plan in the same shape as the root document.

The JSON reporter recursively lays out nested execution graphs before cache checks. Each nested plan keeps dependsOn scoped to its own tasks, so the same task ID can appear at multiple levels. Cache status is applied by walking a path through nested plans rather than only top-level task/step indices.

The run-plan JSON Schema splits steps into CommandStep vs NestedStep, and the changelog documents the new behavior. E2E fixtures replace the former “not supported yet” error case with snapshots for multi-level nested plans, forwarded args, and nested vp run --dry-run (still shown as a spawn step, not expanded).

^{Reviewed by Cursor Bugbot for commit bc48a17. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

fspy benchmark

linux

dynamic/launch             change  +0.80%  [ -7.57% ..  +8.53%]  overhead   +61.18%
dynamic/access             change  +0.29%  [ -5.94% .. +19.74%]  overhead   +20.76%
dynamic/access-relative    change  +0.26%  [ -8.61% ..  +8.10%]  overhead   +73.19%
dynamic/access-contended   change  +4.77%  [ -9.21% .. +21.41%]  overhead    +4.58%
static/launch              change  +0.24%  [ -6.58% ..  +7.94%]  overhead  +196.57%
static/access              change  +0.35%  [ -4.34% ..  +3.38%]  overhead +1481.58%
static/access-relative     change  +0.37%  [ -2.09% ..  +1.89%]  overhead +1977.80%
static/access-contended    change  +0.35%  [ -1.18% ..  +2.43%]  overhead +2065.44%

macos

dynamic/launch             change  -0.27%  [ -4.75% ..  +5.72%]  overhead  +233.84%
dynamic/access             change  +0.36%  [ -2.43% ..  +4.48%]  overhead    +5.03%
dynamic/access-relative    change  -0.43%  [ -5.67% ..  +9.34%]  overhead  +253.12%
dynamic/access-contended   change  +1.67%  [ -8.19% .. +12.15%]  overhead    +2.67%

windows

dynamic/launch             change  -0.08%  [ -3.70% ..  +2.88%]  overhead   +32.40%
dynamic/access             change  +0.00%  [ -1.02% ..  +0.83%]  overhead    +1.47%
dynamic/access-relative    change  +0.00%  [ -0.61% ..  +0.82%]  overhead    +2.05%
dynamic/access-contended   change  -0.20%  [ -2.41% ..  +0.39%]  overhead    +1.76%

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 480cb60. Configure here.}