[FEATURE] `/status` Command: Multi-Agent Plan Progress & Cross-Session Project State

Resolved 💬 3 comments Opened Feb 22, 2026 by adminimb Closed Apr 9, 2026

Preflight Checklist

  • [x] I have searched existing requests and this feature hasn't been requested yet
  • [x] This is a single feature request (not multiple features)

Problem Statement

## Problem Statement

When orchestrating complex multi-step projects with Claude Code:

  1. In-Session Blindness: During parallel task execution (e.g., 3 agents working on different

modules simultaneously), there is no built-in way to query overall progress. Users must manually
check task lists or memory files, interrupting workflow.

  1. Cross-Session Reconstruction: When returning to a project after hours or days, users must

manually parse .claude/memory/ files, review git history, and read plan documents to understand
which phases are complete, which services are in progress, and what blockers exist.

Both scenarios are friction points for long-running projects with multiple agents and
interdependent tasks.

Proposed Solution

Proposed Solution

Add a /status slash command with two layers:

### Layer 1: In-Session Plan Progress
Display a real-time, single-screen summary during active task execution:
╔═══════════════════════════════════════════════╗
║ PARALLEL PLAN PROGRESS ║
╠═══════════════════════════════════════════════╣
║ Agent A (Service Core) [████████ ] 80% ║
║ Agent B (Migrations/DB) [██████ ] 60% ║
║ Agent C (Tests/Handlers) [██ ] 20% ║
║ ║
║ ACTIVE TASKS: ║
║ • A1: Implement endpoints → in_progress ║
║ • B2: Write migrations → blocked by: A1 ║
║ • C1: Unit tests → pending ║
║ ║
║ OVERALL: 7 / 15 tasks (46%) ║
║ EST. COMPLETION: ~2h 30m ║
╚═══════════════════════════════════════════════╝

### Layer 2: Cross-Session State
On startup or upon request, read persistent .claude/plan-state.json (written by previous
sessions) and render:
╔═══════════════════════════════════════════════╗
║ PROJECT STATE (Last active: 2 days ago) ║
╠═══════════════════════════════════════════════╣
║ Service A Phase 1 [DONE] 18 tests ✅ ║
║ Service B Phase 2 [IN PROGRESS] 40% code ║
║ Service C Phase 1 [DONE] migration ✅ ║
║ Integration Tests [BLOCKED] waiting B ║
║ ║
║ LAST CHECKPOINT: ║
║ • Commit: 7ea8c29 (2 days ago) ║
║ • Branch: feature/auth-refresh ║
║ • Open Blockers: 2 ║
╚═══════════════════════════════════════════════╝

Implementation Notes

  • Parse active TaskList (Layer 1) and persistent .claude/plan-state.json (Layer 2)
  • Render to single screen (max 20 lines) — use progress bars and sparse formatting
  • Optional --full flag to expand blockers and recent commits
  • Integrate with existing task tracking system (no new storage layer required for Layer 1)

Alternative Solutions

Alternatives Considered

  1. Pull from git log only: Cannot show in-session task progress; misses real-time agent state
  2. Persistent task database: Adds infrastructure; .claude/ JSON is simpler and git-aware
  3. Interactive dashboard: Out of scope for CLI; single-screen snapshot is sufficient

Priority

Medium - Would be very helpful

Feature Category

CLI commands and flags

Use Case Example

_No response_

Additional Context

_No response_

View original on GitHub ↗

This issue has 3 comments on GitHub. Read the full discussion on GitHub ↗