Minimal Operations Runbook

This is the standard operational checklist for running a HardKAS deployment safely. It relies strictly on observed evidence rather than assumptions.

1. Verify the Workspace

Ensure you are in the correct HardKAS workspace and that the environment is sound.

hardkas env
hardkas doctor

Expected Evidence: WORKSPACE_READY and no fatal errors in the doctor report.

2. Verify the Node (RPC Health)

Before generating any transactions, prove that the target RPC node is alive, synced, and compatible.

hardkas rpc health

Expected Evidence: Node is reachable, isSynced: true, and the version supports the intended features (e.g., >= 2.1.0 for Toccata).

3. Verify Account Authority

Ensure the account executing the transaction has sufficient spendable balance.

hardkas accounts balance deployer_account

Expected Evidence: Balance exceeds the required deployment amount + estimated fee + compute budget.

4. Execution & Artifact Generation

Construct the intent and verify the generated artifact. Do not blindly submit.

hardkas tx plan ... --out plan.json
hardkas artifact inspect plan.json

Expected Evidence: The JSON fields perfectly match your intended recipients, amounts, and compute_budget.

5. Submission & Receipt Verification

Submit the signed transaction and extract the receipt.

hardkas tx send signed_tx.json
hardkas artifact explain receipt.json

Expected Evidence: The command returns a valid txid, and the explanation confirms the network accepted the payload.

6. Procedural Fallback (If Step 5 Fails)

If the transaction is rejected or times out:

1. DO NOT run destructive commands (rm -rf, reset).
2. Run hardkas tx wait <txid> to check if it's stuck in the mempool.
3. Check the failure taxonomy (RPC_SCHEMA_ERROR, FEE_CONVERGENCE_ERROR, etc.) and follow the Troubleshooting Guide.