Test writing, debugging and quality checks.
Adds a full-stack runner workflow with evidence for local or Daytona runs.
Creates, runs and fixes evals that check agent decisions and prompt steering.
Verifies a delivered feature on the real product and writes an evidence report.
Builds agent evaluation systems with metrics, graders, regression tasks, and significance checks.
Detects web performance regressions by comparing page timings against a baseline.
Tests and benchmarks LLM agents for behavior, capability, reliability, and monitoring.
Adds a runner protocol evaluation with its roster, assertions and report fixture.
Automates a real browser for AI agents to test web apps and pages.
Tests code changes from the git diff and reports pass or fail with evidence.
Checks async backend code for event loop blocking and keeps runtime tests in place.
Builds evaluation datasets and runs DeepEval tests for agents, RAG, and LLM apps.
Benchmarks CodeGraph retrieval by comparing agent runs with and without it.
Designs LLM-as-judge systems, pairwise comparisons, rubrics, and bias controls.
Drives a real browser for signed-in sites, forms, screenshots, and web QA.
Drives a real browser to open pages, click flows, take screenshots and check errors.
Drives a real browser to inspect pages, read console and network, and fill forms.
Creates an isolated named Herdr session to reproduce issues without touching the default.
Finds slow, flaky or redundant tests and picks tests that give fast, useful feedback.
Testing workflow and quality standards for writing and running tests.
Sets up repeatable automated accessibility checks, with manual testing still needed.
Test authoring guidance
Test a PR or branch end to end with Docker Compose and API calls.
FastGPT CI workflow 双轨同步。当用户修改或新增 .github/workflows/ 下的 GitHub Actions workflow...
Smoke tests the Agent Builder branch end to end in a scaffolded project.
Use when the user wants to add, write, debug, or configure a Crush hook, gating or...
Runs Android device tests and writes acceptance reports with full screenshots.
Drives a local Electron app over CDP to snapshot, evaluate and smoke-test its UI.
Automates a web task in a local browser and saves screenshots and logs as evidence.
Audits an Anthropic cookbook notebook against a review rubric.
Shows which Xcode test targets were skipped or ran and why test selection changed.
Adds latency or connection failures to outgoing traffic of a mirrord session.
Replays scripted MCP tool calls on an isolated server and checks the rendered cards.
Build, refactor, and test LangBot platform adapters for the Event-Based Agents...
Drives a local OmniRush Electron app over CDP for UI smoke tests.
Tests REST and GraphQL APIs, authentication and failures with curl.
Drives an app in a pty, records a video and opens a checked summary page.
Use this skill to run a four-panel adversarial advisory review on any pull request that...
Optimize the description prompts an AI agent reads to learn its built-in tools (the...
Use kagent agents and standalone sandboxes.
Ground rules for working in the Tika codebase, git policy, Maven wrapper/repo...
Use when a task needs a hidden Linux desktop running in the background or a...
Expert guidance for systematic backtesting of trading strategies.
Provides an advanced financial modeling suite with DCF analysis, sensitivity testing...
Use to extend an existing agent project with memory, app integration, VPC, multi-agent...
Configure, extend, debug, or contribute to PicoClaw itself.
Choose proportional checks for OpenClaw changes and diagnose failing tests.
Authoring playbook for building agents that write, edit, review, or refactor code.
Build, debug, or review Cloudflare Agents SDK applications using the agents package.
