@microsoftOpen source projects and samples from Microsoft
Open source projects and samples from Microsoft
microsoft.comCurated from public sources. This profile has not been claimed.
Multi-model council review for diffs, pull requests, and risky changes — CRITICAL after non-trivial development. USE FOR: "review a diff or pull request", "bug hunt in recent edits", "multiple models inspect the same change independently", "cross-review or debate findings between models", "review after any non-trivial development phase". DO NOT USE FOR: "plan a new implementation", "write code without a review".
Multi-model council planning for implementation and architecture decisions. USE FOR: "plan with council", "multi-model plan", "different perspectives on an approach", "multiple models propose plans independently", "debate approaches or compare strategies", "cross-plan a synthesis". DO NOT USE FOR: "review a diff or pull request", "answer a quick factual question".
Set, inspect, or control a durable goal — the agent keeps iterating across turns until the stop condition is met.
Give the agent a durable objective with a verifiable stopping condition, then keep iterating across turns until that condition is met. Use when the user says 'set a goal', 'follow a goal', '/goal …', 'keep working until …', or asks for a long-running task with a clear end state (migrations, large refactors, retry-until-green loops, experiments).
Set up the Inbox agent with recommended VS Code settings and gh CLI configuration
Search GitHub issues and PRs assigned to you, needing triage, or matching custom queries
Open a GitHub notification URL in the browser
Read and write agent storage — memory and rules with environment-aware fallback
Mark all GitHub notifications as read
Manage subscription for a GitHub notification thread
Fetch GitHub notifications using the gh CLI
Install the GitHub CLI (gh) if not already installed
Get details of a specific GitHub notification thread
Dismiss GitHub notification
Add a reaction to a GitHub issue or pull request
Debug runtime-dependent bugs by generating competing hypotheses, adding temporary structured-log probes, collecting local evidence, and applying only an evidence-backed fix. Use for intermittent or timing-sensitive bugs, races, frontend/backend interaction failures, runtime-state issues, repeated speculative-fix failures, or explicit requests to instrument and collect evidence. Not for breakpoints, stepping, or variable inspection.
Write and update component explorer fixtures for UI work — fixtures, screenshots, visual testing. USE FOR: "create a fixture for a component", "add fixture groups or variants", "screenshots or visual testing of UI", "adding or changing UI in a component explorer project". DO NOT USE FOR: "set up or install the component explorer", "write a unit test with a test runner".
Set up and integrate the Component Explorer into a Vite based project. USE FOR: "set up component explorer", "add component explorer to a Vite app", "install component explorer packages", "configure the component explorer Vite plugin". DO NOT USE FOR: "write a fixture for a component", "set up a different test framework".
Lightweight setup — install Component Explorer packages and add the Vite plugin, with MCP server for AI agent integration. USE FOR: "lightweight or minimal component explorer setup", "install the packages and add the Vite plugin", "component explorer without the CLI daemon", "MCP server for AI agent integration". DO NOT USE FOR: "full setup with VS Code tasks and launch config", "write a fixture for a component".
Full setup of the Component Explorer including CLI, MCP server, VS Code tasks and launch config. USE FOR: "complete component explorer setup with the CLI", "configure the component explorer MCP server", "add VS Code tasks for the component explorer", "add a launch config for the component explorer". DO NOT USE FOR: "minimal or plugin only component explorer setup", "write a fixture for a component".
Analyze VS Code rolling build health on Azure DevOps. Use when: the rolling build is red right now, you need a report for the last 100 builds, you need to identify the commit range that broke the build, you are on build champ duty, or you need a build health report.
Create and manage banned AST patterns that prevent code constructs in edits. USE FOR: "ban a code pattern", "add a tree-sitter lint rule", "prevent specific syntax in edits", "validate a BANNED_AST.md rule". DO NOT USE FOR: "explain how existing code works", "answer general programming questions".
VS Code settings precedence rules and common pitfalls. Essential for any code that reads or writes settings. Covers getConfiguration scope, inspect() vs get(), and multi-workspace handling.
Run smoke tests to verify extension functionality in a real VS Code environment. Use this when checking if basic features work after changes.
1,370 published Skills
1370 Skills · 102 repositories