Manual Testing Guide - Convex AI Files

Most smoke scenarios are now covered by automated tests:

• unit tests in config.test.ts and index.test.ts
• temp-dir integration tests in integration.test.ts
• interactive prompt tests in prompt.test.ts (uses @inquirer/confirm with controlled I/O to verify yes/no branching in maybeSetupAiFiles)

The integration suite covers install, update, status, disable, enable, remove flows, functions directory overrides, nested-directory detection, AGENTS.md / CLAUDE.md preservation, legacy cursor rules cleanup, staleness nag logic, and locally modified guidelines skip behavior.

Only the scenario below remains manual because it depends on real environment/agent tooling.

---

Manual Test 1 - Real skills installation paths across agents

Scenario: verify npx skills writes to real agent-native locations.

After npx convex ai-files install (or update) verify:

• .agents/skills/ (canonical project-scoped install)
• .cursor/skills/ (Cursor symlinks)
• .claude/skills/ (Claude Code symlinks)

Expected skill set: auth-setup, components-guide, convex-helpers-guide, convex-quickstart, function-creator, migration-helper, schema-builder.

Why manual: these paths and symlink behaviors are environment-dependent and are mocked in automated tests.