发现 Skills

Skill Gateway — Rule all skills. PRIORITY: Consult this skill BEFORE any other skill for ANY task-oriented request. Skill Gateway is the router that decides which skill(s) to use. Trigger for: ANY file creation (docs, PDFs, slides, spreadsheets), ANY coding/building task, ANY design/creative request, ANY data analysis or automation, ANY "how do I", "can Claude", "is there a skill for", "help me with" question, ANY vague or ambiguous request. Also trigger when user asks "which skill", "find a skill", wants to automate something, or describes a problem/workflow. ONLY skip for pure conversational chat (greetings, opinions, general knowledge) with no task involved. If ANY task or deliverable is implied, Skill Gateway goes first. Responds in the user's language. Does NOT do work itself — routes to the right skill, repo, or workaround.

Capture frames or clips from RTSP/ONVIF cameras

Control Eight Sleep pods - status, temperature, alarms, schedules

<!-- AUTO-GENERATED by export-skills.py — DO NOT EDIT -->

Auto-review pipeline — reads the full CEO, design, and eng review skills from disk and runs them sequentially with auto-decisions using 6 decision principles. Surfaces taste decisions (close approaches, borderline scope, codex disagreements) at a final approval gate. One command, fully reviewed plan out. Use when asked to "auto review", "autoplan", "run all reviews", "review this plan automatically", or "make the decisions for me". Proactively suggest when the user has a plan file and wants to run the full review gauntlet without answering 15-30 intermediate questions.

Fast headless browser for QA testing and site dogfooding. Navigate any URL, interact with elements, verify page state, diff before/after actions, take annotated screenshots, check responsive layouts, test forms and uploads, handle dialogs, and assert element states. ~100ms per command. Use when you need to test a feature, verify a deployment, dogfood a user flow, or file a bug with evidence. Use when asked to "open in browser", "test the site", "take a screenshot", or "dogfood this".

Post-deploy canary monitoring. Watches the live app for console errors, performance regressions, and page failures using the headless browse daemon. Takes periodic screenshots, compares against pre-deploy baselines, and alerts on anomalies. Use when: "monitor deploy", "canary", "post-deploy check", "watch production", "verify deploy".

CEO/founder-mode plan review. Rethink the problem, find the 10‑star product, challenge premises, expand scope when it creates a better product. Four modes: SCOPE EXPANSION (dream big), SELECTIVE EXPANSION (hold scope + cherry-pick expansions), HOLD SCOPE (maximum rigor), SCOPE REDUCTION (strip to essentials). Use when asked to "think bigger", "expand scope", "strategy review", "rethink this", or "is this ambitious enough". Proactively suggest when the user is questioning scope or ambition of a plan, or when the plan feels like it could be thinking bigger.

Systematically QA test a web application and fix bugs found. Runs QA testing, then iteratively fixes bugs in source code, committing each fix atomically and re-verifying. Use when asked to "qa", "QA", "test this site", "find bugs", "test and fix", or "fix what's broken". Proactively suggest when the user says a feature is ready for testing or asks "does this work?". Three tiers: Quick (critical/high only), Standard (+ medium), Exhaustive (+ cosmetic). Produces before/after health scores, fix evidence, and a ship-readiness summary. For report-only mode, use /qa-only.

Import cookies from your real browser (Comet, Chrome, Arc, Brave, Edge) into the headless browser session. Opens an interactive picker UI where you select which cookie domains to import. Use before QA testing authenticated pages. Use when asked to "import cookies", "login to the site", or "authenticate the browser".

Full safety mode: destructive command warnings + directory-scoped edits. Combines /careful (warns before rm -rf, DROP TABLE, force-push, etc.) with /freeze (blocks edits outside a specified directory). Use for maximum safety when touching prod or debugging live systems. Use when asked to “guard mode”, “full safety”, “lock it down”, or “maximum safety”.

A comprehensive verification system for Claude Code sessions.

Automated skill for keeping AI knowledge bases current with latest model versions, framework updates, and best practices

Fetch the latest AI/ML news, open-source projects, social media discussions, and KOL perspectives. Supports sophisticated filtering by categories and sources.

When the user needs marketing ideas, inspiration, or strategies for their SaaS or software product. Also use when the user asks for 'marketing ideas,' 'growth ideas,' 'how to market,' 'marketing strategies,' 'marketing tactics,' 'ways to promote,' or 'ideas to grow.' This skill provides 139 proven marketing approaches organized by category.

Use this when code conflicts occur during Git merge or rebase. It is especially suited for scenarios where AI automatic conflict resolution often makes mistakes (picking the wrong side, losing refactorings, reverting to older code) and where semantic analysis and logical validation are required. It also applies when a rebase stops across multiple rounds and conflict files must be aggregated across rounds. Trigger words: resolve conflicts, handle conflicts, git-conflict-resolve, resolve merge conflicts, resolve rebase conflicts, conflict resolve, conflict resolution.

Used when a Git repository needs to release a new version. Covers scenarios such as inconsistent tag naming conventions, varying primary branch names (master/main/develop), conflicts when merging a release branch into the mainline, and MR/PR branches containing unrelated files. Suitable for platforms like GitLab, GitHub, and Gitea, and supports single-repository and multi-repository setups. Trigger words: release, tag, publish version, version release, release process, release workflow, git-release, multi-repo release.

Provides file paths to language-specific extension files for the code-testing pipeline. Call this skill to discover available extension guidance files (e.g., dotnet.md for .NET, cpp.md for C++). Do not use directly — invoked by code-testing agents and skills that need language-specific references.

Guide for optimizing MSBuild incremental builds. Only activate in MSBuild/.NET build contexts. USE FOR: builds that are slower than expected on subsequent runs, cases where "nothing changed but it rebuilds anyway", diagnosing why targets re-execute unnecessarily, and fixing broken no-op builds. Covers eight common causes: missing Inputs/Outputs on custom targets, volatile properties in output paths (e.g., timestamps or GUIDs), file writes outside tracked Outputs, missing FileWrites registration, changing globs, and Visual Studio Fast Up-to-Date Check (FUTDC) issues. Key diagnostic: inspect binlog messages for the equivalent of "Building target completely" versus "Skipping target". DO NOT USE FOR: first-time build slowness (use build-perf-baseline), parallelism issues (use build-parallelism), evaluation-phase slowness (use eval-performance), or non-MSBuild build systems. INVOKES: binlog MCP server tools (overview, search, target details); falls back to dotnet build /bl and binlog replay/analysis with diagnostic verbosity.

Migrates .NET test projects from xUnit.net v2 to xUnit.net v3. USE FOR: upgrading xunit to xunit.v3. DO NOT USE FOR: migrating between test frameworks (MSTest/NUnit to xUnit.net), migrating from VSTest to Microsoft.Testing.Platform (use migrate-vstest-to-mtp). For xUnit v3 MTP filter syntax (--filter-class, --filter-trait, --filter-query), also load migrate-vstest-to-mtp.

Build robust backend services with Rust using Clean Architecture. Use when creating APIs, microservices, or server applications with domain-driven design, repository pattern, and async operations.

Structures Git workflow practices. Use when making any code change. Use when committing, branching, resolving conflicts, or when you need to organize work across multiple parallel streams.

Extracts what the user actually wants instead of what they think they should want. Achieves this through a one-question-at-a-time interview until ~95% confidence about the underlying intent. Use when an ask is underspecified ("build me X" without "for whom" or "why now"), when the user explicitly invokes it ("interview me", "grill me", "are we sure?", "stress-test my thinking"), or when you catch yourself silently filling in ambiguous requirements before any plan, spec, or code exists.

Use when the user wants to create a shareable UI mockup, prototype, landing page, dashboard, slide deck, or microsite in Claude Design, or mentions Claude Design, design-system inheritance, brand-on prototypes, exports, or troubleshooting.