waza

Skillgrade scaffolds an eval.yaml for a SKILL.md directory, runs repeated tasks, and measures whether an agent discovers and applies the skill as intended. It supports Gemini, Claude, Codex, ACP, OpenCode, and custom command agents, with deterministic or LLM-rubric graders and CLI or browser reports. Node.js 20+ is required, while Docker, API credentials, or a local command depend on the selected executor.

AgentSight observes AI agents at the system level, correlating process trees, shell commands, file operations, network destinations, prompts, responses, tool intent, and model token data. It works outside closed-source CLIs without an SDK, proxy, or vendor integration, and offers live views, SQLite session reports, and offline pprof profiles. It requires Linux eBPF support and elevated probe privileges, and its capture scope includes potentially sensitive prompts and responses.

Fetch a Databricks job by name, run ID, or workspace URL, then download and decompress driver and executor logs from Unity Catalog Volumes and merge them chronologically with source labels. The CLI supports focused filters and JSONL output. It requires a configured Databricks CLI and does not currently support S3 log destinations.

Use prebuilt CLI verbs to inspect HTML, DOM state, console output, network traffic, screenshots, and scripted interactions against a live deployment. Rehearse can retain per-host authentication and scope Cloudflare Access credentials to the protected origin, while emitting stable JSON for CI, cron, or agent runners. It is Chromium-only and targets diagnostics and repeatable health flows rather than load testing or general synthetic monitoring.