# Plan: Cluster Diagnosis and Script Enhancement (`diagnose_and_enhance`)

## Phase 1: Enhanced Diagnostics (Consul) [x] [checkpoint: a686c5b]
- [x] Task: Update `consul_client.py` to fetch detailed health check output
    - [x] Write tests for fetching `Output` field from Consul checks
    - [x] Implement logic to extract and store the `Output` (error message)
- [x] Task: Update aggregator and formatter to display Consul errors
    - [x] Update aggregation logic to include `consul_error`
    - [x] Update table formatter to indicate an error (maybe a flag or color)
    - [x] Add a "Diagnostics" section to the output to print full error details
- [x] Task: Conductor - User Manual Verification 'Phase 1: Enhanced Diagnostics (Consul)' (Protocol in workflow.md)

## Phase 2: Nomad Integration and Logs [x] [checkpoint: 6d77729]
- [x] Task: Implement `nomad_client.py` wrapper
    - [x] Write tests for `get_allocation_logs`, `get_node_status`, and `restart_allocation` (mocking subprocess)
    - [x] Implement `subprocess.run(["nomad", ...])` logic to fetch logs and restart allocations
- [x] Task: Integrate Nomad logs into diagnosis
    - [x] Update aggregator to call Nomad client for critical nodes
    - [x] Update "Diagnostics" section to display the last 20 lines of stderr
- [x] Task: Conductor - User Manual Verification 'Phase 2: Nomad Integration and Logs' (Protocol in workflow.md)

## Phase 3: Advanced LiteFS Status [ ]
- [x] Task: Implement `litefs_status` via `nomad alloc exec`
    - [x] Write tests for executing remote commands via Nomad
    - [x] Update `litefs_client.py` to fallback to `nomad alloc exec` if HTTP fails
    - [x] Parse `litefs status` output (text/json) to extract uptime and replication lag
- [x] Task: Final Polish and Diagnosis Run
    - [x] Ensure all pieces work together
    - [x] Run the script to diagnose `odroid8`
- [~] Task: Conductor - User Manual Verification 'Phase 3: Advanced LiteFS Status' (Protocol in workflow.md)