agentsclimarketplace

Engineering

98,276 rows, by stacks then stars

  • Injective mcp servers

    ChiJian28/Touchline/.agents/skills/injective-mcp-servers Skill

    0★ repo

    Set up and run Injective MCP servers. Needed by multiple Injective skills which perform MCP tool calls.

  • Engineering cybernetics governance

    Springwonder/engineering-cybernetics-governance Skill

    0★

    Use Qian Xuesen's Engineering Cybernetics as the primary framework for turning work evidence into stable memory, reusable Skills, experiments, and daily feedback control. Use for end-of-day reviews, experience distillation, system improvement, and deciding what belongs in memory versus Skills.

  • Bounded review loop

    takeshijuan/bounded-review-loop/skills/bounded-review-loop Skill

    0★ repo

    Run budget-controlled, evidence-based review and optional repair for pull requests, branches or working-tree diffs, and implementation plan files. Use when asked for multi-angle review, review and fix, repeated review until clean, economy/standard/strict review, or review-only without edits. Keep review fanout and repair loops bounded, route only to models exposed by the live runtime, distinguish blockers from advisory suggestions, re-review targeted surfaces, and report local verification separately from CI, merge, deployment, and production state. Do not use this skill to merge, deploy, publish, approve, or destructively clean up unless those actions are separately authorized.

  • Add skill

    Paldom/stealth-browser-skills/.claude/skills/add-skill Skill

    0★ repo

    Authors or improves an Agent Skill in this repo via the eval-first workflow - scope one purpose, write trigger evals, draft SKILL.md, validate, iterate. Use when the user asks to create, add, write, refine, or fix a skill ("add a skill for X", "improve triggering of Y"). Not for scaffolding a new repository or editing infrastructure (hooks, CI, the validator/scripts, docs).

    The same file is in 11 other repositories

  • Ngrx developer

    kamilfurtak/ngrx-developer/skills/ngrx-developer Skill

    0★ repo

    Design, implement, review, test, debug, and migrate Angular state management with NgRx. Use for @ngrx/store, @ngrx/effects, @ngrx/signals and SignalStore, @ngrx/entity, @ngrx/component-store, @ngrx/router-store, @ngrx/data, Store DevTools, NgRx operators, ESLint rules, schematics, runtime checks, selectors, reducers, actions, effects, entity adapters, and NgRx version migrations.

  • Azpg ai in database

    lestermarch/postgres-ai-experts/skills/azpg-ai-in-database Skill

    no license0★ repo

    Call large language models and Azure AI services from inside Azure Database for PostgreSQL Flexible Server using the azure_ai extension — generate vector embeddings (azure_openai.create_embeddings), run LLM text generation (azure_ai.generate), and use Azure Cognitive Services for sentiment, language detection, key-phrase extraction, entity recognition, and summarization (azure_cognitive.*), all in SQL. Use this skill whenever the task involves in-database AI, calling Azure OpenAI or a Language service from SQL, embedding a column of text, enriching or summarizing rows with an LLM, configuring azure_ai.set_setting endpoints/keys, or bulk-generating embeddings for a table — even when the user just says "add embeddings to my products table" or "summarize these reviews in the database". For designing the vector index/RAG retrieval side hand off to azpg-pgvector-rag.

  • Coordinate github repositories

    TechSpokes/skill-github-repositories-coordination/skills/coordinate-github-repositories Skill

    0★ repo

    Coordinate GitHub repositories across personal accounts and organizations while preserving goals, authority, privacy, workspace roles, and existing workflows. Use for access diagnosis, inventory, findability, portfolio understanding, repository routing, cross-repository work, practice reuse, functionality and knowledge reuse, tool comparison, initial orientation, installation and updates of this skill, or an explicit language quality review of coordination output generated by this skill. Supports software, documentation, writing, research, data, operations, and mixed work. Do not use as the primary workflow for routine work in one known repository or blanket destructive execution inferred only from inactivity; require lifecycle evidence.

  • Scaffold

    UntidyOctopus35/scaffold Skill

    0★

    Structure-first wireframe exploration through a phased interview. Establishes a plain-language screen brief, hierarchy, and constraints before generating four hypothesis-driven low-fidelity directions, running a comparative critique, and producing a final annotated wireframe brief. Use whenever the designer wants to work out a screen or flow's structure before designing it — "help me wireframe this," "how should this screen be laid out," "explore directions for," "what should go on this screen," "lo-fi concepts," "plan the IA for this flow," or any request to plan a screen's content, hierarchy, layout, or step structure before opening a design tool. Trigger even without the word "wireframe." Supports Quick Start — if the designer pastes an existing brief, PRD, or notes, extract answers from it and ask only about gaps. Do NOT use to critique an existing or finished design, to conduct or replace user research, or for visual styling, color, or type decisions.

  • Manuscript review revision

    Jameslxr/manuscript-review-revision-skill/manuscript-review-revision Skill

    0★ repo

    Run a platform-independent, journal-aware, review-first workflow for scientific manuscripts, including target-journal confirmation or Top-5 journal recommendation, live official-author-guideline research, at least five independent reviewer agents, scientific and statistical review, claim-reference verification, reviewer-response drafting, revision, formal DOCX/PDF manuscript formatting, and fail-closed release auditing. Use in Codex, Claude Code, or another compatible Agent Skills host when the user asks to review, revise, polish, format, submit, retarget, recommend journals for, verify references in, or respond to reviewers about a biomedical or scientific manuscript. Always establish the exact target journal before full review; never revise or cosmetically polish before the independent review gate is complete and the user authorizes revision.

  • Architecture governance

    Cartograf666/alex-ter-vibeskills/skills/architecture-governance Skill

    0★ repo

    Bootstrap, document, enforce, and review a project's architecture through layer responsibilities, dependency direction, architecture manifests, legacy baselines, Change Envelopes, cross-layer approval policy, and deterministic architecture checks. Use when starting a project, auditing an existing codebase, planning cross-layer work, preventing architecture erosion, reviewing structural changes, or deciding whether implementation may expand into another layer.

  • Add skill

    Paldom/promptimize/.claude/skills/add-skill Skill

    0★ repo

    Authors or improves an Agent Skill in this repo via the eval-first workflow - scope one purpose, write trigger evals, draft SKILL.md, validate, iterate. Use when the user asks to create, add, write, refine, or fix a skill ("add a skill for X", "improve triggering of Y"). Not for scaffolding a new repository or editing infrastructure (hooks, CI, the validator/scripts, docs).

    The same file is in 11 other repositories

  • Flask monolith refactor

    prithvi020397/agent-refactor-skills/skills/flask-monolith-refactor Skill

    0★ repo

    Use this skill whenever the user asks to refactor, split, restructure, or clean up a single-file Flask (or similar WSGI) backend — one large app.py with many @app.route definitions, module-level state, and no blueprints or package structure. Covers adding failure observability first, removing duplicate routes, characterization tests, extracting pure helpers and services, and converting routes to blueprints — all WITHOUT changing behavior, response shapes, or route contracts, and without touching the frontend.

  • Route github agent decisions

    Mimirs402/bountyverdict-mcp-skill/skills/route-github-agent-decisions Skill

    0★ repo

    Use for: why did a GitHub Actions run fail (root cause); retry a failed workflow; is a GitHub bounty open; which bounty to work on; audit AGENTS.md; will MCP tools/list changes break clients.

  • Independent qa challenge

    ndelanhese/skills/skills/independent-qa-challenge Skill

    0★ repo

    Act as an independent senior QA challenger for feature tasks, product changes, pull requests, bug fixes, and test plans. Use when Codex must discover unplanned scenarios, edge cases, regression risks, missing automated tests, invalid assumptions, or acceptance gaps. The QA mission is to try to invalidate the task with strong evidence, while staying fair, scope-aware, reproducible, and constructive.

  • Jebi agent

    kyutomatte/jebi_agent/skill/jebi-agent Skill

    0★ repo

    Create a clean successor Codex chat for the current project by generating a concise handoff file and using Codex App Server thread/start. Use when the user invokes $jebi, $jebi-agent, JEBI AGENT, 제비, or asks in Korean to move/hand off/continue the current Codex work in a new chat: 제비로 옮겨주세요, 제비로 이사갈게요, 제비한테 넘겨줘, 제비로 인수인계해줘, 제비 출발, 새 채팅으로 제비 보내줘, JEBI로 새 채팅 만들어줘. Do not use for ordinary questions about swallows, images, filenames/code containing jebi, requests to explain JEBI AGENT, or when the user says not to move yet.

  • Repo atlas

    vivekmaru/repo-atlas Skill

    0★

    Create polished repository onboarding documentation as static HTML. Use when a user asks an agent to explore a repo, explain architecture, produce diagrams, render existing Markdown/spec docs into browser-friendly pages, build a glossary of acronyms/domain terms, summarize historical decisions from docs such as .kiro/, or make a new-engineer repo atlas/code map.

  • Shared obsidian brain

    aigeekops/shared-obsidian-brain Skill

    no license0★

    把任意 AI agent 经 MCP 接入 Obsidian vault,搭建本地优先的「第二大脑」 (Karpathy LLM Wiki 范式)。触发词:搭第二大脑 / second brain / LLM Wiki / 接入 Obsidian / 知识库。

  • Bundle optimization

    dgabreuu/react-skill-pack/skills/bundle-optimization Skill

    0★ repo

    Use when analyzing Vite/React bundle composition, initial versus asynchronous transfer, and evidence-based code splitting. Trigger when a production build has a measurable loading or build regression.

  • Tracker audit

    alexclymo/claude-tracker-skills/skills/tracker-audit Skill

    0★ repo

    Use when the user wants to audit the task tracker, clean up the task docs, or says the docs feel stale, out of date, or contradictory — triggers like "audit the task tracker", "clean up the task docs", "the docs feel stale", "reconcile the tracker", "health check the tracker", "do the tracker docs still match reality". Sweeps the whole repo — tracker/, CLAUDE.md, architecture and navigation docs, and code comments — for stale and contradictory claims, reports them ranked by how likely each is to cause a wrong agent decision, and fixes them only on explicit approval. Manages project task *documents* on disk, not Claude Code's in-session task tools (TaskCreate/TaskList) — those are scoped to one conversation and never persist; this skill audits the durable, git-tracked record instead.

  • Skill

    murphy977/3d-eyes/skill Skill

    0★ repo

    Use when the user reports a 3D model or scene looking wrong (black/invisible/gigantic/untextured model, holes, weird shading), asks to check or QA a 3D file (GLB, GLTF, OBJ, PLY, STL, .splat), works with Gaussian splats, or generates 3D assets with AI and wants them verified. Requires the 3d-eyes MCP server.

  • Prompt rail

    yuanGao0816/prompt-rail Skill

    no license0★

    Optimize prompts with a measured train/test dual-gate loop that resists memorizing eval cases. Use when the user wants to write or improve a prompt, tune a system prompt, build prompt evals with holdout, prevent prompt overfitting/背题, run prompt-rail, or iterate intent-classification prompts with separate train and test suites. Fork of prompt-smith plus anti-overfit rails.

  • Session analytics

    shihchengwei-lab/session-analytics Skill

    0★

    Weekly self-improvement loop for AI coding agents, plus ad-hoc usage analytics. Each agent analyzes its OWN tool's local session logs over the last 7 days (Claude Code via /insights artifacts, Codex via ~/.codex/sessions, others via their logs), reports success rates, friction patterns, stalled sessions, tool/token usage — and in a weekly review proposes evidence-tied workflow improvements: a small capped rules block in the tool's config (diff shown, user approves, block self-refactors instead of accumulating), mechanical harness fixes (hooks/env/permissions) preferred over prose rules, and skill/config hygiene both directions (hot paths reinforced, installed-but-never-invoked skills flagged for disabling; rule evidence re-validated across model changes). Trigger on "weekly review", "improve my workflow", "分析我的 session/使用紀錄", "which sessions failed", "my success rate", "compare my experiment runs", "哪些 skill 沒在用/停用沒用的 skill", or bare invocation (= 7-day overview). Reads local data only. Not for regenerating the official /insights report, or deep-reading one full transcript (expensive — see Cost rules).

  • Skill

    RyanKershawWhittle/agentic-splunk-upgrade-orchestrator/skill Skill

    0★ repo

    Splunk Enterprise on-prem upgrade skill for the forwarding/management tier - deployment servers, agent managers, intermediate forwarders, heavy forwarders (incl. DB Connect hosts). Full lifecycle - version-path planning, compatibility gates, pre-checks, per-component runbooks, post-checks, soak, rollback. TRIGGER when the user plans or executes a Splunk version upgrade, patches a forwarder/DS/HF, asks about upgrade order, KV store migration, DB Connect upgrade impact, or rollback. SKIP for app deployment (use splunk skill / splunk-deploy-engineer) and for Splunk Cloud stack upgrades (Splunk-managed).

  • Align system

    mikeqwe/system-reality-alignment-plugin/plugins/system-reality-alignment/skills/align-system Skill

    0★ repo

    Design, review, plan, repair, implement, operate, or evaluate software and socio-technical systems so observations, data, decisions, actions, and outcomes remain traceable and aligned with reality. Use for architecture and system design, data-quality or state-model problems, observability and feedback loops, incident remediation, migrations, domain contracts, invariants, lineage, reconciliation, decision auditability, adaptive operations, and controlled evaluation of an intervention or this plugin. Do not use for isolated cosmetic edits or local syntax changes with no system-behavior implications.

  • Skill press polish

    proluct/skill-press/skills/skill-press-polish Skill

    0★ repo

    Diagnose and improve an EXISTING Claude Code skill with a diagnose, fix, re-diagnose pass, whether skill-press built it or not. Use when the user says "polish this skill", "improve my X skill", "this skill keeps misfiring or confusing the agent", "clean up this SKILL.md", or after a skill-press run offers a polish pass. Not for creating new skills (use skill-press) and not for plugin-managed skills, which cannot be edited in place.

  • Writing claude design prompts

    CodingCossack/writing-claude-design-prompts Skill

    0★

    Use when turning product notes, briefs, specifications, or mixed mission packs into prompts for Claude Opus, Fable, or another design-capable Claude model whose primary result is a visual artefact, frontend, redesign, prototype, design system, or image-to-code implementation.

  • Agent handoff

    sonu07-star/agent-handoff-skill/skills/agent-handoff Skill

    0★ repo

    Create, validate, and resume structured handoff documents that let another AI agent continue work without hidden conversation context. Use when pausing or transferring unfinished work, approaching a context limit, switching agents or sessions, reporting a blocker, recovering after interruption, resuming from HANDOFF.md, or auditing a handoff for evidence, actionable next steps, secret leakage, and unsupported completion claims.

  • Auto tinker automate

    NolanCassidy/auto-tinker/skills/auto-tinker-automate Skill

    0★ repo

    Configure, inspect, or revise Auto-Tinker scheduler policy and exact task prompts without creating a second state system. Use when changing cadence, daily limits, scheduled-task scope, pause behavior, or scheduler handoffs; this skill configures automation but does not perform the scheduled workload or mutate GitHub.

  • Doc steward

    stone16/stometa-skills/skills/doc-steward Skill

    0★ repo

    Steward agent-facing repository documentation against a tiered house standard. Use when auditing or grading AGENTS.md, CLAUDE.md, SKILL.md, .claude/rules, ADRs, or DESIGN.md for structure, drift, frontmatter, or broken routing; explicitly previewing or applying gated low-risk fixes after an audit; or explicitly capturing one documentation lesson. Not for prose copy-editing, application-code review, product-interface design, or generic formatting.

  • Build it out

    SubliminalCoding/build-it-out/skills/build-it-out Skill

    0★ repo

    Run a hardened coordinator-worker-reviewer build board for substantial implementation, hardening, migration, or decompose-build-review-iterate work. Use when a user says "build it out", "use the workflow", "run the board/swarm", requests a hardening pass, or wants independent orchestration and review. Uses a read-mostly Claude coordinator, exact-pane Codex workers, deterministic Git/scope/test gates, and a fresh capability-restricted Claude reviewer with bounded repairs. Do not use for a quick conversational answer or a trivial single-file edit.

  • Optimize claude fable 5 prompts

    IchenDEV/claude-fable-5-prompt-optimizer/plugins/claude-fable-5-prompt-optimizer/skills/optimize-claude-fable-5-prompts Skill

    0★ repo

    Clarify, audit, and rewrite rough or existing prompts for Claude Fable 5 using Anthropic's official prompting guidance. Use when a user asks to optimize, improve, migrate, debug, or design a prompt for Claude Fable 5, including requests phrased as Fable or Fabel prompt optimization, Fable/Fabel 提示词优化, or when an underspecified idea needs guided clarification before a copy-ready prompt. Ask only for missing information that would materially change the result; otherwise preserve intent and produce a lean, context-rich prompt with explicit outcomes, boundaries, evidence, output requirements, and completion criteria.

    The same file is in 1 other repository

  • Spicetify live test

    yusufaf/spicetify-agent-skills/skills/spicetify-live-test Skill

    0★ repo

    Use when testing, debugging, or verifying a Spicetify extension change in the running Spotify desktop client — reading its console, running JS against Spicetify Platform APIs, inspecting or mutating extension localStorage, hot-reloading after an edit, or confirming a fix actually works in the client rather than only passing a syntax check. NOT for Spotify Web API work (playback/playlists via OAuth), which is a different layer.

  • Prompt first workflow

    YMIR777/vibecoding-prompt-workflow/skills/prompt-first-workflow Skill

    no license0★ repo

    Apply first-principles task framing with a bounded read-only preflight before substantive work. Use for projects, changes, debugging, reviews, research, writing, design, and other tasks where goals, context, constraints, approval boundaries, acceptance criteria, or output requirements need to be explicit. Show an executable task definition before substantive work on complex, open-ended, state-changing, or decision-bearing requests; do not force ceremonial rewrites for simple, already-clear answers or low-risk actions.

  • Autonomous learning

    Chulinuwu/skillloom/skills/autonomous-learning Skill

    0★ repo

    Review completed Claude Code, Codex, or compatible agent work for durable declarative knowledge and reusable procedural knowledge. In Skillloom Hermes mode, curate one bounded second-brain outcome only after a checkpointed meaningful delta and policy-promote only validated safe skill candidates. Use when explicitly invoked, when a Stop hook requests learning review, when an invoked host completes non-trivial work, or when repeated corrections reveal a stable workflow.

  • Backend development

    danielleit241/hs-skills/skills/backend-development Skill

    0★ repo

    Build robust backend systems with modern technologies (Node.js, Python, Go, Rust), frameworks (NestJS, FastAPI, Django), databases (PostgreSQL, MongoDB, Redis), APIs (REST, GraphQL, gRPC), authentication (OAuth 2.1, JWT), testing strategies, security best practices (OWASP Top 10), performance optimization, scalability patterns (microservices, caching, sharding), DevOps practices (Docker, Kubernetes, CI/CD), and monitoring. Use when designing APIs, implementing authentication, optimizing database queries, setting up CI/CD pipelines, handling security vulnerabilities, building microservices, or developing production-ready backend systems.

  • Agent skill stack

    neilchen2000-pixel/agent-skill-stack/skills/agent-skill-stack Skill

    0★ repo

    Find, evaluate, assemble, and runtime-activate the smallest compatible set of AI Agent Skills for an end-to-end natural-language goal. Use as the primary orchestrator when a request spans multiple replaceable capabilities or asks for a sustained, complete, or end-to-end outcome—even when the user never says “Skill”—and use domain Skills only as downstream specialists. Also use for multi-Skill selection, installed-Skill audits, conflict checks, low recall, indirect helpers such as humanizers or compliance checks, and project routing with dependency closure. Search local Skills, registries, GitHub, and OpenCLI; compare adoption, verified fit, safety, and overlap; block incomplete handoffs. Do not use for a single bounded action or locating one known/common Skill; use that domain Skill or the generic find-skills workflow.

  • Skill readme workflow

    liush2yuxjtu/skill-readme-workflow Skill

    0★

    启发式编排:将一个 skill 产品化为 README 和 GitHub landing page。

  • Electron development

    Nick-800/skills/skills/electron-development Skill

    no license0★ repo

    High-security and performance-optimized Electron development workflows. Use whenever building, reviewing, extending, or auditing an Electron desktop app — main process, renderer, preload scripts, contextBridge, IPC (ipcMain/ipcRenderer), BrowserWindow config, or packaging with electron-builder/electron-forge. Trigger on any mention of Electron, a desktop app built with Electron, contextIsolation, sandbox mode, preload scripts, IPC bridges, or requests to scaffold/secure/optimize an Electron + TypeScript/React/Vite app. Also trigger when auditing Electron code for security holes (nodeIntegration enabled, disabled contextIsolation, the remote module, missing CSP, unrestricted navigation/window.open) or performance issues (blocking the main process, oversized IPC payloads, listener/memory leaks). Apply the security and performance defaults automatically even if not explicitly requested — they are the non-negotiable baseline, and violating code should be flagged and corrected, not silently written.

  • Exampleboard card workflow

    samuelstefancic/ContextRift/examples/synthetic-project/.agents/skills/exampleboard-card-workflow Skill

    0★ repo

    Guide changes and audits of ExampleBoard card lifecycle transitions, guards, event history, idempotent replay, and lifecycle regression tests. Use for work whose expected behavior depends on docs/card-lifecycle.md. Do not use for copy edits, formatting, generic quality checks, unrelated unit tests, or inventing missing business rules.

  • Frontend behavior preserving refactor

    SHIHAOZOU/ai-frontend-engineering-skills/skills/frontend-behavior-preserving-refactor Skill

    0★ repo

    Improve frontend structure, types, boundaries, readability, or performance without changing intended behavior. Use for component extraction, state simplification, API cleanup, dependency reduction, or architecture migration.

  • Empty litter

    ai-meow/felix-skills/skills/empty-litter Skill

    0★ repo

    Cleaning session for the repository. Scans for bad practices and architectural friction — dead code, duplication, shallow modules, inconsistent patterns, config drift, untested seams — then fixes what the user picks. Use when the user says empty-litter, wants to clean the repo, remove tech debt, or improve the architecture.

  • Clear agent communication

    HustleCoding/clear-agent-communication/skill/clear-agent-communication Skill

    0★ repo

    Explain an AI agent's multi-step work clearly, selectively, and in plain language. Use when investigating problems, changing files or data, running tools, making consequential decisions, encountering blockers or uncertainty, or reporting verification and completion. Communicate observable actions and evidence without exposing private chain-of-thought or narrating every operation.

  • Github project scout

    tmnfootcy/github-project-scout/skills/github-project-scout Skill

    0★ repo

    Use this skill when the user asks to find, compare, audit, review, or recommend GitHub repositories, Codex skills, AI agent frameworks, n8n workflows, automation tools, open-source projects, or role/skill additions that must fit the current project. Prioritize this skill for skill scouting, GitHub project review, and project intelligence work; do not use it for ordinary coding tasks.

  • Sharpemu nid swarm

    Acelogic/sharpemu-agentic-toolkit/skills/sharpemu-nid-swarm Skill

    0★ repo

    Run evidence-backed SharpEmu PS5 NID work from unresolved-import traces through KawaiiDRA reverse engineering, subsystem clustering, isolated worktree implementation, integration, and multi-game regression. Use when an AI coding agent is asked to triage, identify, reverse engineer, implement, batch, delegate, or regression-test SharpEmu NIDs; compare NID coverage with another emulator or PR; or build a parallel NID work queue from logs, CSV catalogs, Aerolib, firmware, Kyty, or game traces.

  • Caveman compress

    arkaigrowth/agent-skills/caveman-compress Skill

    0★ repo

    Compress natural language memory files (CLAUDE.md, todos, preferences) into caveman format to save input tokens. Preserves all technical substance, code, URLs, and structure. Compressed version overwrites the original file. Human-readable backup saved as FILE.original.md. Trigger: /caveman:compress <filepath> or "compress memory file"

  • Agent memory architecture

    wu1chenghui/hermes-math-template/skills/agent-memory-architecture Skill

    no license0★ repo

    Design persistent, tiered memory systems for LLM agents — abstraction-compression pipelines, selective forgetting, hierarchical retrieval, and container-persistence engineering.

  • Azure easy auth

    ericsche/CodingSkills/plugins/codingskills/skills/azure-easy-auth Skill

    0★ repo

    Configure Microsoft Entra ID (Azure AD) authentication for Azure App Service Easy Auth and Azure Static Web Apps. Use when setting up sign-in, protecting APIs, adding Google/Microsoft personal accounts, or debugging 403 "allowed applications" and client-principal issues.

  • Harness engineering

    LiuHHHYuuu/agent-skills/skills/harness-engineering Skill

    0★ repo

    Use whenever the user wants to design, improve, or audit agent-first engineering systems, self-improving Codex workflows, autonomous PR/review loops, vibe-coding guardrails, new systematic projects, greenfield project scaffolding, project knowledge bases, golden principles, cleanup agents, architecture constraints, readable code style, function comments/docstrings, eval loops, durable logs, human judgment allocation, or human approval boundaries. Trigger on phrases like harness engineering, self-improving agents, agent-first, vibe coding, new project, start a serious project, build a system, AGENTS.md, golden principles, drift, garbage collection, cleanup agent, context management, code style for agents, reviewable code, agent-to-agent review, or Building self-improving tax agents with Codex.

  • Flame graph skills

    celeroncoder/flame-graph-skills Skill

    0★

    Flame-graph style page/site audit: capture Chrome Performance traces + network, rank hot paths (main-thread, JS, LCP, bundles), map to concrete fixes, optional re-measure. Triggers: /flame-audit, flame graph audit, profile this page, find perf hot paths, optimize website from flamegraph, CPU/main-thread bottlenecks, audit this site like a flame graph, where is the hot path on this page.

  • Delegate to claude

    wakamiya0315/delegate-to-claude/skill/delegate-to-claude Skill

    0★ repo

    Use this skill by default before the supervising agent directly edits repository code when work adds or updates tests, likely changes two or more files, creates a non-trivial module, investigates a failure, performs a focused refactor or code review, or likely needs more than three repository tool calls. Delegate the bounded execution slice to a supervised Claude Code Sonnet worker, using a concise prompt when possible. Direct editing is only for an unambiguous low-risk change of about ten lines or less in one file with no tests or investigation. Keep architecture, ambiguous product decisions, security-sensitive work, secrets, external side effects, commit/push/deploy, and final approval with the supervisor.

  • Swiftdata testing

    akshaypimprikar/ios-swiftdata-testing-agent-skill/swiftdata-testing Skill

    0★ repo

    Writes and reviews SwiftData test code — in-memory ModelContainer fixtures, mock repository patterns, @ModelActor test isolation, Decimal money-value assertions, and content-hash dedup testing. Use when writing or reviewing unit/integration tests for a SwiftData-backed app.

  • Architect maintainable apps

    AE86-Victory/webapp-engineering-skills/skills/architect-maintainable-apps Skill

    0★ repo

    Design maintainable application architecture before and during implementation. Use when starting a web, mobile, desktop, or backend app; adding a substantial feature; choosing MVC, MVVM, unidirectional flow, factories, dependency injection, or singleton lifetimes; defining folder structure and module boundaries; or preventing UI, state, business logic, and infrastructure from becoming coupled.

  • Repository hygiene audit

    multiplai-ai/repository-hygiene-audit/skills/repository-hygiene-audit Skill

    0★ repo

    Audit one or more Git repositories for exact duplicates, suspicious near-duplicates, misplaced context, inconsistent project structure, dirty worktrees, nested repositories, raw-data risks, credential-like files, and other structural drift. Use for recurring repository cleanup reviews, context-folder governance, pre-refactor inventories, or scheduled read-only hygiene reports.

  • Monodraw llm

    p404/monodraw-llm/.claude/skills/monodraw-llm Skill

    0★ repo

    Turn natural-language descriptions, ASCII/text diagrams, screenshots, and existing `.monopic` or `.monojson` files into native editable Monodraw documents. Create, reconstruct, inspect, validate, and regression-test complex diagrams with separate resizable frames, nested groups, attached or free-endpoint connectors, arrows, shadows, z-order, and pinned connector labels. Use whenever an LLM should sketch a diagram in ASCII and export it to Monodraw, recreate a screenshot or legacy v3/current v5 document 1:1, or diagnose Monodraw serialization and editability.

  • Claude to codex migrator

    craxelfn/claude-to-codex-migrator/skills/claude-to-codex-migrator Skill

    no license0★ repo

    Migrate Claude skills, plugins, assistant packages, ZIP archives, pasted files, repository trees, or implementation documentation into clean Codex Skills or plugins. Use when Codex must inventory a source package, choose Skill versus plugin architecture, preserve behavior, rewrite platform-specific instructions, validate MCP or hooks, remove Claude branding from the distributable output, and produce migration and cleanup reports.

  • Llm privacy

    patrikherak/llm-privacy Skill

    0★

    Reversible PII tokenization for working with sensitive data. Masks personal data (emails, phones, IBANs, cards, national IDs, tagged names/addresses) behind stable placeholder tokens BEFORE text enters model context, and restores the real values in the final output — deterministic code, no ML, no network. Use when reading, querying, or processing data that contains (or may contain) personal information of real people — customer records, database query results, CRM exports, support tickets, log files — so the model never sees the actual PII but the user still gets a complete answer.

  • Social agent public workflows

    lniass/social-skills/skills/social-agent-public-workflows Skill

    0★ repo

    Write, review, approve, and schedule Facebook posts for a business through the hosted Social Agent service. Use this whenever someone asks for a social media post or caption, wants to see posts already prepared for them, wants to approve or schedule one, wants to connect their Facebook Page, or wants recurring posting set up. Post copy comes from the hosted service and is never written locally. Covers first-time setup through secure Handled verification as well as returning users who already have a project.

  • System audit review

    MaiDuy708/system-audit-review Skill

    0★

    Use when auditing or reviewing a system, repository, service, configuration, data pipeline, security boundary, automation, or operational workflow, especially a large, mixed, risky, or forensic target.

  • Youtube transcript

    flamefalcon7/youtube-transcript Skill

    0★

    Transcribe YouTube videos locally with Whisper into agent-ready transcripts — no captions, no cloud. Per-channel/domain vocabulary memory makes it more accurate every time the user corrects a term. Use when given a YouTube URL to transcribe, process, summarize, or sync into a knowledge base.

  • Cohort method

    TJCurnutte/cohort-method/.agents/skills/cohort-method Skill

    0★ repo

    Run multi-agent cohort reviews, creative panels, UX swarms, and cross-QC passes (e.g. "10 agents / 5 cohorts"). Spawns parallel persona or scoped agents, synthesizes consensus, and soft-merges findings into actionable next ships without reshuffling live work. Use when the user asks for a cohort review, creative panel, UX swarm, taste panel, multi-agent QC, or phrases like "spin up N agents" / "N agents / M cohorts".