GITHUB AGENT SKILLS / 技能收录
给你的 AI 助手
装上专业技能
我们从 GitHub 持续收录、审核并分类优质的 Agent Skills,按职业与使用场景整理,复制一行命令即可安装到 Claude Code、Cursor 等智能体。
按职业浏览 Agent Skills
为不同岗位挑选的高频技能
按使用场景浏览 Agent Skills
从你要完成的任务出发
按创作者浏览 Agent Skills
按 GitHub Stars 排序的优秀创作者
Microsoft
@microsoftopenclaw
@openclawAnthropic
@anthropicsOpenAI
@openaiThe Apache Software Foundation
@apacheGitHub
@githubAutoGPT
@Significant-GravitasGarry Tan
@garrytan全部 Skills 482556
reframe-transcribe
Use when transcribing speech from audio/video to text with WORD-LEVEL timing (for captions, karaoke subtitles, clip-boundary snapping, or diarization). The proven local pipeline is faster-whisper (large-v3-turbo on GPU / small on CPU) for the transcript, then a SEPARATE ctc-forced-aligner second.
reframe-scrollshot-markdown
Reconstruct a vertically scrolled screen recording into an aligned long screenshot, evidence-preserving OCR, and Markdown through the Reframe app. Use for scroll-capture videos, recorded long screenshots, chat/page/table scrolling captures, or requests to verify a Reframe scrollshot evidence bundle. Do not use for ordinary 16:9-to-9:16 subject reframing.
refactor-equivalence
Use when a change CLAIMS to be a pure/deterministic refactor - extract-method, inline, rewrite a helper, "no functional change" - and you want an EXTERNAL proof that old and new are equivalent. CrossHair diffbehavior uses Z3 to find an input where old_fn(x) != new_fn(x); a Hypothesis differential test is the fallback. Scoped to PURE, DETERMINISTIC functions; not for non-Python. Advisory only.
red
Offensive operations package — full tier tables (FEW/MEDIUM/HEAVY/FULL/MAX/SOTA) for wiring the estate's offensive arsenal with exact enable AND disable commands, including the agent-integrity/jailbreak vector library. Child of red-blue-purple.
recursive-decision-ledger
Use when the user asks for repeated rollouts, marked decision processes, high-dimensional search, stochastic optimization, local-optima exploration, ensemble comparison, or recursive reasoning with a visible evidence trail.
recover-from-false-positive
暂无简介
receiving-code-review
Use when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementation
recall-workflow-agent
Use when you need the exact input PROMPT or final OUTPUT of a specific subagent from a past Workflow / ultracode fan-out run - reviewing or recovering what one fanned-out agent (a lens, a verifier, the synthesizer) actually received or returned after the run scrolled off screen or the session compacted. Identify it by a distinctive phrase, the workflow name, or by listing the run's agents.
reboot-safe-ops-continuation
Use when a reboot is required in the middle of migration or repair work and you need exact pre-reboot state capture, staged-artifact inventory, and a resumable checklist for the next session.
re-triage-preflight
Run a fast, hash-cached reverse-engineering TRIAGE preflight on an unknown binary BEFORE deep analysis — folds file-format detection, Shannon-entropy packer triage (works with zero external tools), Detect-It-Easy (diec), FLOSS auto string-deobfuscation, and capa capabilities into ONE structured TRIAGE block, cached by SHA-256 so re-runs are instant. Use when you are about to reverse a native binary (PE/ELF/Mach-O/DEX/.so), triage an unknown/suspicious file, check whether a sample is packed/obfuscated, or want a cheap first-pass before opening IDA/Ghidra/radare2. Advisory only — never blocks, degrades gracefully when a tool is missing. Inert until an RE context. Triton dynamic-taint and VulnHuntr/objdiff+m2c are documented opt-in follow-ons; the revula mega-bundle is rejected.
re-ensure
Bring up the reverse-engineering MCP backends (IDA / Ghidra) on demand so mcp__ida-pro-mcp / mcp__ghidra-mcp calls "just work" without manually opening the GUI apps. Use when asked to start/ensure the RE stack, or before a batch of RE tool calls. Cold-starts ONLY the requested backend, prefers an already-running GUI IDA on :13337, and reports that Binary Ninja (Personal) cannot run headless.
ralphinho-rfc-pipeline
RFC-driven multi-agent DAG execution pattern with quality gates, merge queues, and work unit orchestration.
ralph-orchestrator-setup
Use when installing, updating, validating, or troubleshooting the local Ralph Orchestrator CLI, its MCP server registration, or the mirrored Ralph skills for Codex, Claude, and Gemini.
ralph-loop
Run, monitor, resume, merge, and debug Ralph loops. Use this skill whenever the user asks to operate `ralph run` or `ralph loops`, inspect loop state, recover suspended loops, analyze diagnostics, or unblock merge queue issues.
ralph-hats
Create, inspect, validate, explain, and improve Ralph hat collections. Use this skill whenever the user asks to make or refine a `.ralph/hats/*.yml` workflow, debug hat routing, explain event topology, or tune a multi-hat Ralph run.
radare2
Use when command-line disassembly, debugging, hex editing, or scriptable low-level binary triage is needed without relying on a heavyweight GUI first.
avoid-ai-writing
Use whenever drafting or editing copy, prose, UI text, titles, headings, or subject lines. Review generic AI writing patterns while preserving facts, voice, uncertainty, protected content, and the requested scope.
quality-zero-platform-control-plane
Use when working with `Prekzursil/quality-zero-platform`, enrolling or auditing governed repos, checking strict-zero wrapper workflows, ruleset drift, trusted-runner Codex remediation, or provider/bootstrap state for the personal-account control plane.