Context & AI engineering
25,815 rows, by stacks then stars
1,865 of these skill files have been read, by 1,328 distinct authors, and what they tell an agent counted. What Context & AI engineering authors agree on, and what they forbid
8ddieHu0314/Skill-Lab/tests/fixtures/skills/security/prompt-injection Skill
54★ repoUse when you need to generate a summary of a document.
liqiongyu/lenny_skills_plus/skills/building-with-llms Skill
54★ repoProduce an LLM Build Pack (prompt+tool contract, data/eval plan, architecture+safety, launch checklist). See also: ai-evals (eval only), ai-product-strategy (strategy only).
liqiongyu/lenny_skills_plus/skills/delegating-work Skill
54★ repoCreate a Delegation Pack (brief, decision rights, context handoff, check-in cadence, debrief).
ariffazil/arifOS/io.github.ariffazil/arifos-mcp MCP server
53★ repoConstitutional AI governance server with 5-organ Trinity and enforced floors F1-F13.
vinvcn/addyosmani-agent-skills-zh/skills/context-engineering Skill
52★ repo优化 agent 上下文设置。当开始新会话、agent 输出质量下降、在任务之间切换,或需要为项目配置规则文件和上下文时使用。
vinvcn/addyosmani-agent-skills-zh/skills/using-agent-skills Skill
52★ repo发现并调用 agent skills。用于开始会话,或需要判断当前任务适用哪个 skill 时。这个 meta-skill 负责治理所有其他 skills 的发现与调用方式。
asteroid-belt/skulto/.claude/skills/superplan Skill
51★ repoUse when starting significant features, epics, or complex tasks. Creates multi-phase implementation plans with parallelizable phases, poker estimates, TDD-first acceptance criteria, and quality gates. Detects tech stack from CLAUDE.md/AGENTS.md (bypassing internet research if complete) or via codebase scan.
haabe/mycelium/plugins/mycelium/skills/eval-runner Skill
50★ repoRun eval scenarios to benchmark Mycelium effectiveness. Execute tasks using reflexion loop, validate against success criteria, record metrics.
haabe/mycelium/plugins/mycelium/skills/handoff Skill
50★ repoAI made building cheap. It didn't make deciding cheap. Mycelium is a Claude Code harness that makes your agent run discovery and weigh evidence before it writes code. It earns the right to start. Built for software, courses, AI tools, and services.
haabe/mycelium/plugins/mycelium/skills/prompt-optimizer Skill
50★ repoA/B test CLAUDE.md instruction changes against eval benchmarks. Capture baselines, test variants, compare results.
haabe/mycelium/plugins/mycelium/skills/scaffold-cost-check Skill
50★ repoMeasure Mycelium's own scaffold token cost (CLAUDE.md + engine + harness + canvas + memory) and surface a structured estimate. One-shot audit; pair with /framework-health for trend tracking.
jcottam/agent-resources/skills/productivity/handoff Skill
50★ repoCompact a chat session or task into a curated markdown handoff file under ~/.agents/handoffs/ for continuation in a new agent turn or chat. References existing plans and PRDs by path instead of duplicating them; optional next-session focus via arguments. Use when the user says handoff, compact context, resume handoff, continue session, carry over context, checkpoint, or start fresh with prior work. Not the content-wireframe implementer handoff.
Stitch mcp apply design system
gabelul/stitch-kit/skills/stitch-mcp-apply-design-system Skill
49★ repoApplies a Stitch Design System to existing screens — updates their colors, fonts, and roundness to match the design system's theme. Requires an assetId from list or create operations.
Stitch mcp create design system from design md
gabelul/stitch-kit/skills/stitch-mcp-create-design-system-from-design-md Skill
49★ repoCreates a Stitch Design System from a DESIGN.md that was already uploaded with stitch-mcp-upload-design-md. Second half of a two-step pair — the upload must happen first.
The same file is in 2 other repositories
Stitch mcp create design system
gabelul/stitch-kit/skills/stitch-mcp-create-design-system Skill
49★ repoCreates a reusable Stitch Design System from theme tokens — colors, fonts, roundness. Can be applied to future screens for visual consistency.
gabelul/stitch-kit/skills/stitch-mcp-create-project Skill
49★ repoCreates a new Stitch project container (a design workspace). Call this when starting a new design session to get a projectId for screen generation.
gabelul/stitch-kit/skills/stitch-mcp-delete-project Skill
49★ repoPermanently deletes a Stitch project and all its screens. Destructive — requires explicit user confirmation before calling.
The same file is in 3 other repositories
gabelul/stitch-kit/skills/stitch-mcp-edit-screens Skill
49★ repoEdits existing Stitch screens with text prompts — the iteration tool. Change colors, layout, content, or style without regenerating from scratch. Supports refinement loops via output_components suggestions.
The same file is in 2 other repositories
Stitch mcp generate screen from text
gabelul/stitch-kit/skills/stitch-mcp-generate-screen-from-text Skill
49★ repoGenerates a high-fidelity UI screen or wireframe from a text prompt using Stitch. This is the core text-to-UI generation tool — the heart of the Stitch workflow.
gabelul/stitch-kit/skills/stitch-mcp-generate-variants Skill
49★ repoGenerates design variants of existing Stitch screens using the native variant API. Explore alternative layouts, color schemes, fonts, or content with configurable creativity levels.
The same file is in 2 other repositories
gabelul/stitch-kit/skills/stitch-mcp-get-project Skill
49★ repoRetrieves metadata for a specific Stitch project — title, theme, create/update time. Use to inspect a project before generating new screens. Do NOT use this if you already have a screenId — use stitch-mcp-get-screen instead.
gabelul/stitch-kit/skills/stitch-mcp-get-screen Skill
49★ repoRetrieves full details of a specific Stitch screen — HTML download URL, screenshot URL, dimensions. This is the final step in design retrieval before code conversion.
Stitch mcp list design systems
gabelul/stitch-kit/skills/stitch-mcp-list-design-systems Skill
49★ repoLists all Stitch Design Systems, optionally filtered by project. Returns asset names needed for apply and update operations.
gabelul/stitch-kit/skills/stitch-mcp-list-projects Skill
49★ repoLists all Stitch projects accessible to the user. Use this to find an existing project ID before resuming work or generating new screens in an existing project.
The same file is in 3 other repositories
gabelul/stitch-kit/skills/stitch-mcp-list-screens Skill
49★ repoLists all screens in a Stitch project. Use this after generate_screen_from_text to find the screenId of the newly generated screen, then call stitch-mcp-get-screen to retrieve it.
The same file is in 3 other repositories
Stitch mcp update design system
gabelul/stitch-kit/skills/stitch-mcp-update-design-system Skill
49★ repoUpdates an existing Stitch Design System's theme, tokens, or guidelines. Requires the asset name from create or list operations.
gabelul/stitch-kit/skills/stitch-mcp-upload-design-md Skill
49★ repoUploads a DESIGN.md file to a Stitch project as the first half of turning it into a Stitch Design System. Always paired with stitch-mcp-create-design-system-from-design-md, which must be called immediately after.
The same file is in 2 other repositories
gabelul/stitch-kit/skills/stitch-orchestrator Skill
49★ repoMaster entry point for all Stitch design workflows. Routes from user request → design spec → prompt assembly → screen generation → iteration (edit, variants, design systems) → design system extraction → framework conversion (Next.js, Svelte, HTML, React Native, or SwiftUI) → optional quality pass.
gabelul/stitch-kit/skills/stitch-setup Skill
49★ repoStep-by-step installer for the stitch-kit plugin and Stitch MCP server. Use this when setting up the plugin for the first time, diagnosing connection issues, or helping a new user get Stitch running in Claude Code or Codex CLI.
The same file is in 3 other repositories
gabelul/stitch-kit/skills/stitch-skill-creator Skill
49★ repoMeta-skill for creating new stitch-kit skills. Enforces naming conventions, SKILL.md structure, examples format, and the Design-First SOP. Use when adding new framework support, a new domain-specific prompt architect, or any new capability to the stitch-kit plugin.
The same file is in 3 other repositories
gabelul/stitch-kit/skills/stitch-ued-guide Skill
49★ repoVisual vocabulary, design terminology, and prompt engineering strategy for Stitch. Reference this when you need layout pattern names, aesthetic style terms, color structure formulas, or device guidelines.
gabelul/stitch-kit/skills/stitch-ui-prompt-architect Skill
49★ repoBuilds Stitch-ready prompts via two paths — Path A enhances vague ideas into polished prompts, Path B merges a Design Spec JSON + user request into a structured [Context] [Layout] [Components] prompt.
The same file is in 2 other repositories
ContextJet-ai/awesome-llm-observability/skills/add-llm-guardrails Skill
no license49★ repoUse this to add safety and security guardrails to an LLM/agent app - blocking prompt injection, PII leakage, jailbreaks, toxic output, off-topic responses, or invalid structured output. Trigger on "add guardrails", "prevent prompt injection", "stop PII leaks", "validate the model's output", "make this safe for production", especially for regulated/finance/enterprise use.
ContextJet-ai/awesome-llm-observability/skills/build-eval-dataset Skill
no license49★ repoUse this to build a good evaluation dataset for an LLM app, the part everyone underestimates. Trigger on "make an eval set", "what should I test my LLM on", "I don't have test data for my prompt", "build a golden dataset", or before setting up evals. A great eval set beats a great metric; garbage-in means your evals lie to you.
ContextJet-ai/awesome-llm-observability/skills/collect-user-feedback Skill
no license49★ repoUse this to capture user feedback on LLM outputs (thumbs up/down, edits, corrections, implicit signals) and feed it back into observability and evals. Trigger on "add thumbs up/down", "collect feedback on responses", "how do I know if users like the answers", "improve from real usage", "human feedback loop". Turn real user signal into your best source of eval data.
ContextJet-ai/awesome-llm-observability/skills/compare-llm-models Skill
no license49★ repoUse this to pick or switch the LLM behind a feature, based on evidence instead of hype or the newest release. Trigger on "which model should I use", "is GPT/Claude/Gemini/Llama better for this", "should I switch models", "can a cheaper model do this", "compare models for my use case". Evaluate on YOUR task, not on leaderboards alone.
ContextJet-ai/awesome-llm-observability/skills/debug-agent-from-traces Skill
no license49★ repoUse this to diagnose WHY an LLM agent or chain produced a wrong, empty, slow, or expensive result, by reading its observability trace. Trigger on "my agent gave the wrong answer", "the chain returned nothing", "why is this so slow/expensive", "debug this trace/run", or when a trace tree is available. Walk the span tree systematically instead of guessing.
ContextJet-ai/awesome-llm-observability/skills/estimate-llm-cost Skill
no license49★ repoUse this to estimate what an LLM call or feature will cost, and to compare models on price, before or after shipping. Trigger on "how much will this cost", "estimate my OpenAI/Anthropic bill", "is a cheaper model worth it", "cost of this prompt", "project my LLM spend". Ships a runnable, tested calculator so the numbers are real, not hand-waved.
ContextJet-ai/awesome-llm-observability/skills/eval-driven-development Skill
no license49★ repoUse this to build or change an LLM feature the reliable way, by writing evals first and iterating against them, instead of tweaking prompts by vibes. Trigger on "how do I improve this prompt", "my changes keep breaking other things", "how do I know if this is better", "iterate on my agent", or any prompt/model/RAG change. This is test-driven development for LLMs.
ContextJet-ai/awesome-llm-observability/skills/instrument-llm-observability Skill
no license49★ repoUse this when adding tracing/observability to an LLM or AI-agent application - capturing prompts, tool calls, token usage, latency, and cost per step. Trigger whenever someone wants to "add tracing", "instrument", "monitor", "see what my agent is doing", or debug an LLM app in production. Prefer vendor-neutral OpenTelemetry unless a specific platform is already in use.
ContextJet-ai/awesome-llm-observability/skills/measure-agent-task-success Skill
no license49★ repoUse this to measure whether an AI agent actually completed its task end to end, not just whether individual LLM calls looked fine. Trigger on "is my agent working", "measure agent success rate", "evaluate my agent", "how good is my agent", "agent completion rate", or evaluating a multi-step/tool-using agent. Score the outcome of the whole task, plus the path it took.
ContextJet-ai/awesome-llm-observability/skills/monitor-rag-quality Skill
no license49★ repoUse this to measure and monitor the quality of a RAG (retrieval-augmented generation) pipeline - whether it retrieves the right context and answers faithfully. Trigger on "my RAG gives wrong answers", "is my retrieval any good", "the chatbot makes things up", "evaluate my RAG", "improve RAG accuracy". Diagnose whether the failure is in retrieval or generation - they need different fixes.
ContextJet-ai/awesome-llm-observability/skills/optimize-prompts Skill
no license49★ repoUse this to improve a prompt systematically instead of hand-tweaking it by feel. Trigger on "optimize my prompt", "make this prompt better", "the prompt isn't working well", "auto-tune my prompt", "few-shot example selection", or when prompt quality has plateaued. Optimize against an eval set with a method, and let the numbers pick the winner.
ContextJet-ai/awesome-llm-observability/skills/red-team-llm-app Skill
no license49★ repoUse this to adversarially test an LLM/agent app before attackers do - prompt injection, jailbreaks, data exfiltration, tool misuse, and unsafe output. Trigger on "red team my LLM", "test for prompt injection", "is my agent secure", "jailbreak testing", "security review of my AI app", especially before shipping anything customer-facing or with tools/data access. Test systematically against the known attack classes, not ad-hoc.
ContextJet-ai/awesome-llm-observability/skills/reduce-llm-cost Skill
no license49★ repoUse this to cut the cost of an LLM app using observability data. Trigger on "my OpenAI/Anthropic bill is too high", "reduce token usage", "the app is expensive", "optimize LLM cost", "why am I spending so much on the API". Find the expensive spans first (measure), then apply the cheapest wins. Don't guess - the trace tells you where the money goes.
ContextJet-ai/awesome-llm-observability/skills/trace-multi-agent-system Skill
no license49★ repoUse this to add observability to a multi-agent or agentic system (multiple agents, sub-agents, tool loops, handoffs). Trigger on "trace my agents", "my agent crew is a black box", "which agent failed", "debug my LangGraph/CrewAI/AutoGen/agent workflow", "the agents loop forever". Get a clear span tree across agents and tool calls so you can see who did what.
UnitOneAI/SecuritySkills/skills/ai-security/agent-security Skill
49★ repoReviews AI agent architectures for security risks including permission model design, least-privilege enforcement, human-in-the-loop gate placement, blast radius containment, audit trail completeness, rollback capability, and multi-agent trust boundaries. Auto-invoked when reviewing agentic AI systems where LLMs invoke tools, take autonomous actions, or operate in multi-agent configurations. Produces a structured architecture security assessment mapped to OWASP Agentic AI threats and NIST AI RMF 1.0.
UnitOneAI/SecuritySkills/skills/ai-security/agentic-top-10 Skill
49★ repoReviews agentic AI systems against the OWASP Top 10 security risks for autonomous AI agents. Auto-invoked when reviewing multi-agent architectures, AI agent deployments, or systems where LLMs have tool access and act autonomously. Covers permission models, tool security, memory integrity, trust boundaries, and human oversight. Produces a structured assessment with risk ratings and architectural recommendations.
UnitOneAI/SecuritySkills/skills/ai-security/llm-top-10 Skill
49★ repoReviews LLM-powered applications against the OWASP Top 10 for Large Language Model Applications (2025 edition). Auto-invoked when reviewing code that integrates LLM APIs, builds RAG pipelines, or deploys AI-powered features. Produces a structured findings report mapped to LLM01-LLM10 with severity ratings, CWE mappings, and prioritized remediation guidance.
UnitOneAI/SecuritySkills/skills/ai-security/prompt-injection Skill
49★ repoTests LLM applications for prompt injection vulnerabilities per OWASP LLM01:2025. Covers direct injection (user input manipulating model behavior) and indirect injection (external content containing hidden instructions). Auto-invoked when reviewing LLM applications that process external content, build RAG pipelines, or accept user input that reaches a language model. Produces a test report with categorized findings and defense recommendations.
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/alirezarezvani-skills/skill-security-auditor Skill
48★ repoSecurity audit and vulnerability scanner for AI agent skills before installation. Use when: (1) evaluating a skill from an untrusted source, (2) auditing a skill directory or git repo URL for malicious code, (3) pre-install security gate for Claude Code plugins, OpenClaw skills, or Codex skills, (4) scanning Python scripts for dangerous patterns like os.system, eval, subprocess, network exfiltration, (5) detecting prompt injection in SKILL.md files, (6) checking dependency supply chain risks, (7) verifying file system access stays within skill boundaries. Triggers: "audit this skill", "is this skill safe", "scan skill for security", "check skill before install", "skill security check", "skill vulnerability scan".
The same file is in 1 other repository
Assessing vector and embedding weaknesses
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/assessing-vector-and-embedding-weaknesses Skill
48★ repoTest vector stores for embedding inversion, cross-tenant leakage, and poisoning.
Auditing mcp servers for tool poisoning
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/auditing-mcp-servers-for-tool-poisoning Skill
48★ repoScan Model Context Protocol servers and tool metadata for poisoning, SSRF, and unauthenticated exposure.
Continuous llm red teaming with promptfoo
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/continuous-llm-red-teaming-with-promptfoo Skill
48★ repoWire Promptfoo and DeepTeam into CI/CD for automated regression red-teaming of LLM apps against OWASP LLM Top 10 and OWASP Agentic presets, failing the build when jailbreak or injection vulnerabilities regress.
Detecting ai model prompt injection attacks
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/detecting-ai-model-prompt-injection-attacks Skill
48★ repo'Detects prompt injection attacks targeting LLM-based applications using a multi-layered defense combining regex pattern matching for known attack signatures, heuristic scoring for structural anomalies, and transformer-based classification with DeBERTa models. The detector analyzes user inputs before they reach the LLM, flagging direct injections (system prompt overrides, role-play escapes, instruction hijacking) and indirect injections (encoded payloads, multi-language obfuscation, delimiter-based escapes). Based on the OWASP LLM Top 10 (LLM01:2025 Prompt Injection) and Simon Willison''s prompt injection taxonomy. Activates for requests involving prompt injection detection, LLM input sanitization, AI security scanning, or prompt attack classification.
The same file is in 1 other repository
Detecting indirect prompt injection
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/detecting-indirect-prompt-injection Skill
48★ repoDetect and defend against prompt injection hidden in documents, web pages, and images consumed by an agent.
Extracting config from agent tesla rat
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/extracting-config-from-agent-tesla-rat Skill
48★ repoExtract embedded configuration from Agent Tesla RAT samples including SMTP/FTP/Telegram exfiltration credentials, keylogger settings, and C2 endpoints using .NET decompilation and memory analysis.
The same file is in 1 other repository
Extracting memory artifacts with rekall
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/extracting-memory-artifacts-with-rekall Skill
48★ repo'Uses Rekall memory forensics framework to analyze memory dumps for process hollowing, injected code via VAD anomalies, hidden processes, and rootkit detection. Applies plugins like pslist, psscan, vadinfo, malfind, and dlllist to extract forensic artifacts from Windows memory images. Use during incident response memory analysis.
The same file is in 1 other repository
Implementing llm guardrails for security
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/implementing-llm-guardrails-for-security Skill
48★ repo'Implements input and output validation guardrails for LLM-powered applications to prevent prompt injection, data leakage, toxic content generation, and hallucinated outputs. Builds a security validation pipeline using NVIDIA NeMo Guardrails Colang definitions, custom Python validators for PII detection and content policy enforcement, and the Guardrails AI framework for structured output validation. The guardrails system intercepts both user inputs (blocking injection attempts, stripping PII, enforcing topic boundaries) and model outputs (detecting hallucinations, filtering toxic content, validating JSON schema compliance). Activates for requests involving LLM output validation, AI content filtering, guardrail implementation, or LLM safety enforcement.
The same file is in 1 other repository
Orchestrating llm attacks with pyrit
adriannoes/awesome-agentic-ai/cursor-claude-codex/skills/anthropic-cybersecurity-skills/skills/orchestrating-llm-attacks-with-pyrit Skill
48★ repoBuild multi-turn, Crescendo, and Tree-of-Attacks-with-Pruning (TAP) automated attack chains against conversational LLM agents using Microsoft PyRIT, with adversarial chat and scorer feedback loops.