Skip to main content

This page is intended for AI Coding Agents, not people.

Testing Strategy

Use this workflow after creating or editing any FTD JSON.

Step 1: Static self-check (agent)​

Before returning output, verify:

  • JSON parses successfully.
  • Required keys are present.
  • No extra keys exist.
  • Value types and ranges are correct.
  • script has no obvious syntax issues.
  • If TX is present, RX cleanup exists.
  • If timers are present, cleanup exists.

Step 2: Validator command​

Run the FTD validator with this exact command:

npx --yes --package figaro-tools@latest figaro-ftd-validate <path-to-task.json>

Validation policy:

  • Run validation after creating or modifying any FTD file.
  • If validation fails, fix the JSON and rerun until it passes.
  • Report either success or exact validator errors.

When working inside this repository, also run:

npm run validateScripts

This catches shape errors and common script-safety problems.

Step 4: Manual behavior test plan​

Perform a controlled manual test:

  1. Confirm initial radio state.
  2. Run task once and verify expected CAT side effects.
  3. Cancel mid-flow and verify cleanup executes.
  4. Repeat run to confirm idempotent behavior.
  5. For autorun tasks, reconnect and verify priority/order behavior.

Output note requirement​

After validation, include a short status note such as:

  • Validation passed: figaro-ftd-validate reported 0 errors.

If validation did not pass, include the exact failing messages.