Final Approach

The landing sequence for completing a mission. Runs the Go Poll to get status from all stations.

Usage

/final-approach [issue-number]

What This Does

Final Approach is the end-of-mission workflow:

1. Sync - Fetch and rebase with upstream
2. Check - Run ./scripts/check.sh
3. Go Poll - Review agents report status in reentry mode
4. Landing Doc - Create summary document
5. Commit - Ready for merge

Go Poll

The Go Poll launches review agents (all use Haiku model for speed):

A+ Skip Rule

If an agent gives ALL A+ in round N, skip that agent in round N+1.

Example:

Round 1: Run all 3 agents
  - mission-control: A+  ← Will be skipped
  - telemetry: B
  - flight-director: A

Round 2: Skip mission-control, run telemetry + flight-director
  - telemetry: A
  - flight-director: A+  ← Will be skipped

Round 3: Skip mission-control + flight-director, run only telemetry
  - telemetry: A+

Done: All agents at A or A+

Output Structure

tmp/{ISSUE}/
└── landing/
    └── round-{N}/
        ├── mission-control.md
        ├── telemetry.md
        └── flight-director.md

Grading Scale

Success Criteria

Must achieve A+ from ALL THREE agents before landing.

If any agent gives below A:

1. Fix all identified issues
2. Run ./scripts/check.sh
3. Re-run /final-approach
4. Repeat until all agents report "go"

Instructions

When invoked, you MUST:

1. Determine the issue number from context (git branch, plan file, or ask user)
2. Determine the round number (start at 1, increment for re-reviews)
3. Check previous round grades - skip agents with A+ from prior rounds
4. Create the output directory: tmp/{ISSUE}/landing/round-{N}/
5. Launch agents IN PARALLEL using Task tool with model: "haiku/sonnet"
6. After all agents complete, summarize the grades in a table
7. If any grade is below A +, list the blocking issues that must be fixed

Example agent prompts:

Agent 1 (mission-control):

Go Poll for issue #{ISSUE} - Final Approach (Round {N}).

You are Mission Control reporting status for landing.

Focus: Plan compliance and documentation quality.

1. Find the plan file (check ~/.claude/plans/reovim/ or tmp/)
2. Verify all planned phases are implemented
3. Audit documentation quality (module docs, API docs, guides)
4. Check CHANGELOG.md for completeness

Write your report to: tmp/{ISSUE}/landing/round-{N}/mission-control.md

End with: "Mission Control: GO / NO-GO" and grade (A+/A/B/C/F).

Agent 2 (telemetry):

Go Poll for issue #{ISSUE} - Final Approach (Round {N}).

You are Telemetry reporting status for landing.

Focus: Test coverage and test quality.

1. Inventory all tests (unit and integration)
2. Analyze coverage of happy paths, error paths, edge cases
3. Detect test anti-patterns (flaky, slow, giant tests)
4. Identify missing critical test cases

Write your report to: tmp/{ISSUE}/landing/round-{N}/telemetry.md

End with: "Telemetry: GO / NO-GO" and grade (A+/A/B/C/F).

Agent 3 (flight-director):

Go Poll for issue #{ISSUE} - Final Approach (Round {N}).

You are Flight Director reporting status for landing.

Focus: Code quality and Unix philosophy compliance.

1. Verify Unix philosophy: mechanism vs policy, single responsibility, composability
2. Check layer boundaries (kernel → drivers → modules)
3. Audit unsafe code for SAFETY documentation
4. Run cargo clippy and cargo doc to verify zero warnings
5. Assess complexity and anti-patterns

Write your report to: tmp/{ISSUE}/landing/round-{N}/flight-director.md

End with: "Flight Director: GO / NO-GO" and grade (A+/A/B/C/F).

CRITICAL: Launch agents using the Task tool with:

• subagent_type set to the agent name (mission-control, telemetry, flight-director)
• model: "haiku" for fast, efficient reviews
• model: "sonnet" for comprehensive, precise reviews
• Skip agents that received A+ in previous rounds