Develop visor AI assistants with YAML tests, validation, mocks, and trace debugging.
npx skills add https://github.com/probelabs/visor --skill visor-agent-dev
Please help me install this Agent Skill. Skill: visor-agent-dev Source: https://github.com/probelabs/visor/tree/main/.claude/skills/visor-agent-dev Command: npx skills add https://github.com/probelabs/visor --skill visor-agent-dev
Guides developers through a safe, end-to-end development loop for visor AI assistants, from skill creation to debugging.
Write YAML tests first, then validate and iterate with mocks using visor validate and visor test.
High-intent search queries and answers about installing and using this skill.
Testing AI assistant configurations safely involves writing YAML tests first, validating configurations, and iterating with mocks before graduating to real providers. This structured workflow enforces safety guardrails and trace-based debugging throughout the development loop.
Trace-based debugging for AI assistant development is a mechanism that uses execution traces to validate safety rules and diagnose issues when modifying assistant configurations. It enables iterative testing by graduating configurations from mocked environments to real providers.
Yes, YAML test files and assistant configurations can be validated locally using the visor validate command. This allows developers to check configurations, run tests with mocks using visor test, and ensure safety rules are enforced before production.
The best way to extend AI assistants with new skills and workflows is following a structured end-to-end development loop. This approach requires writing YAML tests first, validating configurations, iterating with mocks, and using trace debugging to graduate changes safely.
AI assistant configuration validation fails when safety rules are violated or configurations lack required elements. Using trace-based debugging and running visor validate helps identify specific issues in the YAML test files before graduating to real providers.
tnezInitialize the .docent directory structure and configuration for new projects.
vnmoorthyMonitor live apps for console errors and performance regressions post-deploy.
anouar1991Intercept and mock API responses during automated tests with Playwright.
botPhUDesign and configure Jest test suites for TypeScript projects.
