lewing

Investigate .NET CI failures using the hlx CLI tool via bash. USE FOR: checking Helix job status, searching build logs, downloading test results, AzDO build timeline analysis — when MCP tools aren't loaded, when the MCP server isn't configured or fails to start, or when scripting with JSON output and jq. DO NOT USE FOR: tasks where Helix/AzDO MCP tools are already available in context (prefer ci-analysis skill when MCP server is loaded). INVOKES: bash (hlx CLI commands with --json output).

{what this skill teaches agents}

Audit pattern for comparing WorkItemSummary (list API) fields against WorkItemDetails (per-item API) after a Helix.Client SDK bump

Domain-layer normalizer + JSON-stable cache key pattern for filter types with server-side defaults and case-insensitive fields.

Strict unknown-parameter rejection for MCP tools: Stage A (UnmappedMemberHandling.Disallow) + Stage B (did-you-mean CallToolFilter).

Audit MCP/CLI tool parameter surface against underlying REST API capabilities to find silently-dropped params.

Test MCP CallToolFilter behavior with real CallToolRequestParams and McpServerTool invocation.

Completion-safe bucketing for Helix work items with nullable exit codes

Analyze CI build and test status from Azure DevOps and Helix for dotnet repository PRs. Use when checking CI status, investigating failures, determining if a PR is ready to merge, or given URLs containing dev.azure.com or helix.dot.net. Also use when asked "why is CI red", "test failures", "retry CI", "rerun tests", "is CI green", "build failed", "checks failing", or "flaky tests". DO NOT USE FOR: investigating stale codeflow PRs or dependency update health, tracing whether a commit has flowed from one repo to another, reviewing code changes for correctness or style.

Measure and reduce the per-session context cost of an MCP server's tool surface. USE FOR: auditing my MCP server tokens, where are my MCP tokens spent, shrink my tool surface, MCP context cost, tools/list cost, why is my MCP server expensive, measuring tool description size, compare MCP versions before/after, per-field MCP token breakdown, MCP token budget, ranking MCP tools by cost. DO NOT USE FOR: designing tool descriptions from scratch (use mcp-server-design), implementing MCP protocol mechanics, benchmarking tool latency or runtime cost (this skill measures static tool-surface tokens, not execution).

Guide MCP server design for agent consumption: tool descriptions, naming, knowledge tools, annotations. USE FOR: writing tool descriptions, naming tools, designing knowledge endpoints, reviewing MCP server design, adding tool annotations. DO NOT USE FOR: building skills that consume MCP tools (use skill-builder), MCP protocol implementation details.

Build, test, and deploy Copilot CLI skills. Use when creating a new skill, improving an existing skill, testing skills with subagents, deploying skills to a repo, or learning skill development patterns.

Institutional knowledge for skill training — frontmatter rules, INVOKES patterns, stop signals, token budgets, anti-patterns, and eval integration. USE FOR: looking up skill-building conventions, understanding training methodology, creating Arena eval requests, checking external reference sources. DO NOT USE FOR: actually training a skill (use SkillTrainer agent), creating new skills from scratch (use skill-builder). INVOKES: no tools — pure knowledge skill.

Guide contributions to this plugin marketplace repository. USE FOR: installing skills, syncing assets, diffing installed vs repo, checking what's out of date, "how do I install a skill", "sync my skills", "what's out of date", "how do I use the CLI". DO NOT USE FOR: building skills from scratch (use skill-builder).

Analyze WebAssembly binaries from .NET WASM builds using wasm-objdump and related tools. USE FOR comparing dotnet.native.wasm across runtime pack versions, verifying SIMD instruction presence, diagnosing execution mode changes, file size forensics. DO NOT USE FOR building WASM apps from scratch or running benchmarks.

Parse and triage performance regression issues auto-filed by the dotnet/performance pipeline. USE FOR categorizing regressions by severity, detecting infrastructure artifacts vs real regressions, identifying data gaps, correlating with dotnet/performance repo changes. DO NOT USE FOR running benchmarks (use runtime-performance) or binary analysis (use wasm-binary-analysis).

{what this skill teaches agents}

Strict unknown-parameter rejection for MCP tools: Stage A (UnmappedMemberHandling.Disallow) + Stage B (did-you-mean CallToolFilter).

Query .NET APIs across NuGet packages, platform libraries, and local files. Search for types, list API surfaces, compare versions, find extension methods and implementors. Use whenever you need to answer questions about .NET library contents.

Query Maestro/BAR dependency flow data using the mstro CLI tool via bash. USE FOR: subscription health checks, build flow tracing, codeflow status, channel discovery, triggering subscription updates — when MCP tools aren't loaded or when scripting with JSON output and jq. Also use when investigating "is this subscription stale", "what's the latest build", "check backflow status". DO NOT USE FOR: tasks where maestro MCP tools are already available in context (prefer flow-analysis or flow-tracing skills when MCP server is loaded). INVOKES: bash (mstro CLI commands with --json output).