Arness Infra Save Plan

Convert a PLAN_PREVIEW_INFRA_*.md plan preview into a structured project directory with infrastructure-specific INTRODUCTION.md, per-phase PHASE_N_PLAN.md files, TASKS.md, report templates, and a PROGRESS_TRACKER.json for execution tracking. This is the bridge between change-plan (which produces the preview) and execute-change (which consumes the structured project).

Pipeline position:

arn-infra-init -> arn-infra-change-spec -> arn-infra-change-plan -> **arn-infra-save-plan** -> arn-infra-execute-change -> arn-infra-review-change -> arn-infra-document-change

Prerequisites

Read ## Arness from the project's CLAUDE.md. If no ## Arness section exists or Arness Infra fields are missing, inform the user: "Arness Infra is not configured for this project yet. Run /arn-infra-wizard to get started — it will set everything up automatically." Do not proceed without it.

Check the Deferred field. If Deferred: yes, inform the user: "Infrastructure is in deferred mode. Plan structuring is not available until infrastructure is fully configured. Run /arn-infra-assess to un-defer." Stop.

Extract:

• Infra plans directory -- where structured plan projects live (e.g., .arness/infra-plans)
• Infra report templates -- which report template set to use (default: default)
• Infra template path -- where to copy report templates for the project (e.g., .arness/infra-templates)
• Infra template version -- current template version for checksum tracking
• Experience level -- derived from user profile. Read ~/.arness/user-profile.yaml (or .claude/arness-profile.local.md if it exists — project override takes precedence). Apply the experience derivation mapping from ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-ensure-config/references/experience-derivation.md. If no profile exists, check for legacy Experience level in ## Arness as fallback.

If Infra plans directory is not configured, stop and instruct the user: "Infra plans directory is not configured. Run /arn-infra-init to set up infrastructure pipeline configuration."

Workflow

Step 1: Locate Plan Preview

Search for PLAN_PREVIEW_INFRA_*.md files in the Infra plans directory.

Glob <infra-plans-dir>/PLAN_PREVIEW_INFRA_*.md

If no preview found: Also check the project root:

Glob PLAN_PREVIEW_INFRA_*.md

If exactly one preview found: Auto-select it and present: "Found plan preview: [filename]. Proceeding to structure this plan."

If multiple previews found: Present the list:

Ask (using AskUserQuestion):

"Which plan would you like to structure?"

Options:

1. [filename1] -- [first line or title]
2. [filename2] -- [first line or title] ...

If still no preview found: Inform the user: "No PLAN_PREVIEW_INFRA_*.md found. Run /arn-infra-change-plan first to generate a plan preview."

Step 2: Extract Plan Metadata

Read the selected plan preview file. Extract:

• Project name -- from the plan title (convert to kebab-case for directory name)
• Phase count -- number of phases defined
• Phase titles -- per-phase title and target environment
• Total resources -- count of resources across all phases
• Blast radius summary -- per-phase blast radius classification
• Environment promotion order -- environments in deployment sequence

Present a summary:

"Plan Summary:

• Project: [project name]
• Phases: [count]
• Resources: [total count]
• Environments: [list in promotion order]
• Blast radius: [summary]

Ask (using AskUserQuestion):

"Proceed with structuring this plan?"

Options:

1. Yes -- Structure the plan into a project directory
2. No -- Cancel

Step 3: Create Project Directory Structure

Run the save_infra_plan.sh script to create the project directory:

bash ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/scripts/save_infra_plan.sh "<project-name>" "<infra-plans-dir>" "<plan-preview-path>"

This creates:

<infra-plans-dir>/<project-name>/
├── SOURCE_PLAN.md          (copy of the original preview)
├── plans/                  (phase plan directory)
└── reports/                (execution report directory)

Step 4: Generate INTRODUCTION.md

Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/references/infra-templates.md for the INTRODUCTION.md template structure.

Generate INTRODUCTION.md in the project directory using the template from infra-templates.md. Populate all sections from the plan preview content:

• Change Overview -- what is changing and why
• Infrastructure Context -- providers, IaC tools, current state summary
• Current State -- existing resources, environment status
• Target State -- desired end state
• Blast Radius Summary -- per-phase classification
• Rollback Strategy -- checkpoint locations, rollback procedures
• Environment Promotion Pipeline -- promotion order, gates between environments
• Cost Budget -- per-phase estimates, total budget
• Security Requirements -- compliance frameworks, access controls

Write to: <infra-plans-dir>/<project-name>/INTRODUCTION.md

Step 5: Generate PHASE_N_PLAN.md Files

Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/references/infra-templates.md for the PHASE_N_PLAN.md template structure.

For each phase in the plan preview, generate a PHASE_N_PLAN.md file in the plans/ directory using the infrastructure-specific template. Each phase plan includes:

• Phase Overview and target environment
• Resources table (create/modify/destroy with resource type, provider, action)
• Blast Radius classification and justification
• Rollback Checkpoint (what to snapshot, restore procedure)
• Security Requirements for this phase
• Cost Estimate (monthly delta)
• Dependencies
• Tasks (implementation task list)
• Acceptance Criteria

Write to: <infra-plans-dir>/<project-name>/plans/PHASE_N_PLAN.md (one per phase)

Step 6: Generate TASKS.md

Generate TASKS.md with the consolidated task list from all phases. Format:

# Tasks

## Phase 1: [title]
- [ ] TASK-P1-001: [description]
- [ ] TASK-P1-002: [description]

## Phase 2: [title]
- [ ] TASK-P2-001: [description]
...

Write to: <infra-plans-dir>/<project-name>/TASKS.md

Step 7: Copy Report Templates

Copy the 3 infrastructure report templates from the plugin to the project's Infra template path:

Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/report-templates/default/INFRA_CHANGE_REPORT_TEMPLATE.json Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/report-templates/default/INFRA_REVIEW_REPORT_TEMPLATE.json Read ${CLAUDE_PLUGIN_ROOT}/skills/arn-infra-save-plan/report-templates/default/INFRA_PROGRESS_REPORT_TEMPLATE.json

Write each template to the Infra template path directory. Generate SHA-256 checksums for each copied template using sha256sum or shasum -a 256 for future update detection.

Store checksums in <infra-template-path>/.checksums.json:

{
  "version": "<infra-template-version>",
  "files": {
    "INFRA_CHANGE_REPORT_TEMPLATE.json": "<sha256>",
    "INFRA_REVIEW_REPORT_TEMPLATE.json": "<sha256>",
    "INFRA_PROGRESS_REPORT_TEMPLATE.json": "<sha256>"
  }
}

Step 8: Generate PROGRESS_TRACKER.json

Generate PROGRESS_TRACKER.json in the project directory following the INFRA_PROGRESS_REPORT_TEMPLATE.json schema. Initialize all phases with not_started status:

{
  "reportType": "infra-progress",
  "projectName": "<project-name>",
  "lastUpdated": "<current-iso-timestamp>",
  "overallStatus": "not_started",
  "totalPhases": <N>,
  "changeSpec": "<path-to-spec>",
  "changePlan": "<path-to-SOURCE_PLAN.md>",
  "phases": [
    {
      "phaseNumber": 1,
      "phaseTitle": "<title>",
      "planFile": "plans/PHASE_1_PLAN.md",
      "environment": "<target-env>",
      "blastRadius": "<classification>",
      "execution": { "status": "not_started", "reportFile": "reports/INFRA_CHANGE_REPORT_PHASE_1.json" },
      "securityGate": { "status": "not_started" },
      "costGate": { "status": "not_started" },
      "verification": { "status": "not_started" },
      "review": { "verdict": "", "reportFile": "" }
    }
  ],
  "notes": []
}

Write to: <infra-plans-dir>/<project-name>/PROGRESS_TRACKER.json

Step 9: Present Summary

Adapt the summary presentation to the user's experience level.

Expert: "Structured project created: <infra-plans-dir>/<project-name>/

• INTRODUCTION.md, [N] phase plans, TASKS.md, PROGRESS_TRACKER.json
• Report templates copied to <infra-template-path>/
• Run /arn-infra-execute-change to begin execution."

Intermediate: Present the file tree with brief descriptions of each artifact and suggest the next step.

Beginner: Present the file tree with explanations of what each file is for and a step-by-step guide for what to do next.

Next steps:

1. Execute the plan -- Run /arn-infra-execute-change to begin phased execution
2. Review a phase plan -- Read any PHASE_N_PLAN.md to review before execution
3. Edit the plan -- Modify any phase plan or INTRODUCTION.md before proceeding

Error Handling

• ## Arness config missing: Suggest running /arn-infra-wizard to get started. Stop.
• No PLAN_PREVIEW_INFRA_*.md found: Suggest running /arn-infra-change-plan to generate a plan preview first. Stop.
• Directory creation fails: Report the error with the path that failed. Check permissions and disk space.
• Script execution fails: Report the script error. Fall back to creating the directory structure manually using mkdir -p.
• Template copy fails: Report which template failed to copy. The project can still function -- templates can be copied manually later.
• Checksum generation fails: Warn: "Could not generate template checksums. Template update detection will not work until checksums are regenerated." Continue without checksums.
• Re-running is safe: Re-running overwrites INTRODUCTION.md, PHASE_N_PLAN.md files, TASKS.md, and PROGRESS_TRACKER.json. Report templates are only re-copied if checksums indicate they have not been customized. If checksums mismatch, ask the user: "Report templates have been customized. Overwrite with the latest versions, or keep your customizations?"