Context & AI engineering
25,610 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
nick3/vibe-design-skills/skills/vibe-design-execute Skill
0★ repoExecute a concrete product or UI design request end to end using the project's approved Vibe Design foundation: clarify the feature, choose or explore structure, create or update the real design/prototype, cover interaction states, evaluate rendered evidence, and iterate to a governed stopping point. Use this skill whenever a user asks an Agent to design, redesign, prototype, or substantially extend a screen, flow, dashboard, app, website, or agentic experience—not merely to author foundation documents. Do not use it to deploy, ship, or rewrite production systems unless the user separately authorizes those actions.
axiom-wang/Axiom_Skills_Created/refine-prompt Skill
no license0★ repoConvert vague product, UI, backend, fullstack, or coding ideas into concise, precise agent implementation prompts by inspecting the current repository or workspace. When requirements are vague, ask Socratic questions to clarify before refining.
tejask0/agent-skills/obsidian-update-agentmd Skill
0★ repoScan the Obsidian vault's current folder/file structure and regenerate CLAUDE.md to reflect it accurately. Use when the user says "update CLAUDE.md", "sync the vault structure", "I've added new notes/folders", "update the agent context", or "run obsidian-update-agentmd". Also run after any bulk note reorganisation.
tejask0/agent-skills/vault-sync Skill
0★ repoReconcile Claude's memory with the user's Obsidian vault, then mirror the result back into CLAUDEMEM.md. Use this whenever the user says "sync my vault", "sync from my notes", "update your memory from my notes", "catch up on my notes", "vault sync", or otherwise asks Claude to refresh what it knows about them from their knowledge base. Reads a fixed set of core vault files (NEVER CLAUDEMEM.md itself), diffs them against current memory, updates ONLY what changed, and then rewrites the changed parts of CLAUDEMEM.md so the agent-context file stays in step with memory. Intended to be run as a periodic ritual (e.g. weekly review), not every day.
hashorva/purrfect-harness/.agents/skills/mcp-hygiene Skill
no license0★ repoHow to use MCP servers (Supabase, browser, GitHub, any connector) without polluting the context window. MUST be used whenever a task involves querying an MCP server, verifying database state, fetching external data mid-task, or when a session is getting long and slow. Also use when deciding whether to reach for MCP at all versus a CLI/script.
Paldom/mcp-server-skills/skills/mcp-server-auth Skill
0★ repoImplements MCP server authorization - OAuth 2.1 resource server role, RFC 9728/8414 discovery, PKCE, RFC 8707 audience validation, scopes, client registration. Use when adding auth, OAuth, tokens, or credentials to an MCP server, or debugging 401/403/discovery failures. Not for general threat hardening, transport wiring, or non-MCP app login.
Paldom/mcp-server-skills/skills/mcp-server-deployment Skill
0★ repoDeploys, operates, and distributes MCP servers - stateless scaling, session externalization, observability, health probes, canary rollouts, MCPB bundles, npm/PyPI packaging, MCP Registry publishing, versioning. Use when the user asks to deploy, host, scale, operate, package, publish, or version an MCP server. Not for writing server code, auth flows, hardening, or evals.
Paldom/mcp-server-skills/skills/mcp-server-design Skill
0★ repoDesigns an MCP server's tool/resource/prompt surface - outcome-oriented consolidation, tool budget, naming, descriptions-as-prompts, flat typed schemas, pagination, agent-readable errors. Use when planning what an MCP server should expose, fixing wrong-tool choices by improving names and descriptions, or turning an API into MCP tools. Not for writing server code, OAuth, or deployment.
Paldom/mcp-server-skills/skills/mcp-server-implementation Skill
0★ repoImplements MCP servers with the official TypeScript or Python SDK - registering tools/resources/prompts, stdio and Streamable HTTP transports, lifecycle, isError vs protocol errors, structured output, progress and cancellation. Use when the user asks to build, create, code, scaffold, or debug an MCP server or its transport. Not for choosing the tool surface, OAuth, hardening, evals, or hosting.
Paldom/mcp-server-skills/skills/mcp-server-security Skill
0★ repoHardens MCP servers against MCP-specific attacks - tool poisoning, rug pulls, prompt injection via tool results, session hijacking, SSRF, command injection - with guardrails-in-code, sandboxing, audit logging, kill switches. Use for securing, hardening, threat-modeling, or security-reviewing an MCP server or vetting a third-party one. Not for OAuth/token wiring or generic appsec scans.
Paldom/mcp-server-skills/skills/mcp-server-testing Skill
0★ repoTests and evaluates MCP servers - Inspector, the official conformance suite, agent evals with verifiable answers, transcript-driven description iteration, tool-selection accuracy, CI regression gates, load tests. Use when the user asks to test, evaluate, verify, or measure an MCP server or its tool selection. Not for fixing server code, security hardening, or generic unit testing.
trugurpala/divan/plugins/core-pack/skills/dispatching-parallel-agents Skill
29 days old0★ repoUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependencies
The same file is in 134 other repositories
trugurpala/divan/plugins/core-pack/skills/subagent-driven-development Skill
29 days old0★ repoUse when executing implementation plans with independent tasks in the current session
The same file is in 124 other repositories
trugurpala/divan/plugins/sadrazam/skills/defterdar Skill
29 days old0★ repoOpen-source workflow layer for Codex, Claude Code and other coding agents: planning, persistent context, explicit execution and verifiable delivery.
trugurpala/divan/plugins/sadrazam/skills/ordu-nizami Skill
29 days old0★ repoOpen-source workflow layer for Codex, Claude Code and other coding agents: planning, persistent context, explicit execution and verifiable delivery.
trugurpala/divan/plugins/sadrazam/skills/vezir-yetistirme Skill
29 days old0★ repoSkill-creation and evaluation coach for the Divan marketplace ("training a new vizier"). Use when the user wants to add, improve, benchmark or test a skill, write a SKILL.md, contribute to the marketplace, compare skill-enabled behavior with a baseline, or says yeni skill yaz, Divan'a vezir ekle, skill nasıl yazılır, skill'i test et, benchmark this skill, evaluate triggering, or create a new skill for Divan. Creates a standards-compliant skill, a bounded eval contract, marketplace wiring and teftiş evidence without inventing performance gains.
trugurpala/divan/plugins/zanaat-pack/skills/claude-api Skill
29 days old0★ repoClaude API and Anthropic SDK reference for model selection, pricing, parameters, streaming, tool use, MCP, agents, caching, token counting and migrations. Use whenever the request names Claude, Anthropic, Opus, Sonnet, Haiku, the anthropic packages or Claude model IDs; when choosing or debugging a Claude API feature; or when an LLM-shaped task names no provider. Before editing, inspect the target for another provider. Do not use this skill when the user or project explicitly targets OpenAI, Gemini, Llama, Mistral, Cohere or Ollama; ask before switching an existing non-Anthropic implementation to Claude. Verify time-sensitive model, price and limit claims against current official documentation.
trugurpala/divan/plugins/zanaat-pack/skills/mcp-builder Skill
29 days old0★ repoGuide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
The same file is in 42 other repositories
TopicalitySummit/summit-decision-assurance/skills/prepare-agent-release-review Skill
no license0★ repoRun a public-safe release review that combines Summit Decision Receipt preflight with governed-output declaration checks. Use before an agent report, recommendation, decision artifact, or consequential output is released. Do not use this workflow as an approval, certification, or substitute for human judgment.
awesome-liuxiao/agent-skill-doctor/benchmarks/public/v1/fixtures/dynamic-eval Skill
no license0★ repoSynthetic mutation with a dynamic-evaluation helper.
awesome-liuxiao/agent-skill-doctor/benchmarks/public/v1/fixtures/prompt-injection Skill
no license0★ repoSynthetic adversarial skill used only inside containment fixtures.
awesome-liuxiao/agent-skill-doctor/benchmarks/public/v1/upstream/anthropic/skill-creator Skill
no license0★ repoCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
The same file is in 66 other repositories
awesome-liuxiao/agent-skill-doctor/benchmarks/public/v1/upstream/openai/skill-creator Skill
no license0★ repoGuide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Codex's capabilities with specialized knowledge, workflows, or tool integrations.
The same file is in 3 other repositories
cristianmoroaica/bountyverdict/skills/preflight-agent-skills Skill
0★ repoAudit public agent SKILL.md bundles with SkillVerdict before installation or execution. Use when evaluating a third-party Codex, Claude Code, Gemini, OpenClaw, or cross-agent skill; checking scripts, credential access, prompt injection, persistence, destructive commands, remote execution, symlinks, submodules, secrets, or undeclared capabilities; or deciding whether a skill should be installed, manually reviewed, or blocked.
cristianmoroaica/bountyverdict/skills/route-github-agent-checks Skill
0★ repoRoute GitHub engineering and MCP catalog questions to the correct BountyVerdict x402 decision API. Use for bounty selection, coding-agent instruction or skill audits, workflow diagnosis, flake retry gates, and tools/list compatibility or security checks before accepting a server change.
ahtishamshahzad/agent_dev_flow/.ai/skills/project-orchestrator Skill
0★ repoUse as the lead skill for any non-trivial request. It drives the full request→approval pipeline (classify, audit, requirements, applications, stack, architecture, phases, tasks) and coordinates specialist skills. It plans and delegates; it does not implement every domain itself, and it stops before implementation until gates are approved.
lzx-Bill/Project-Factory-Core/.agents/skills/copilot-session-retro Skill
0★ repoCopilot Chat Session 专项复盘工具:自动读取当前 Copilot Chat session 的完整对话历史(transcript JSONL), 多维度分析 AI 与用户在 session 中的协作表现,生成结构化复盘报告,并可将改进措施应用到项目配置文件中。 两个模式:review(生成复盘报告)、apply(将报告中的改进措施写入配置文件)。 触发关键词:session review、session retro、copilot retro、session复盘、对话复盘、Copilot回顾。 **注意:本 skill 仅用于 Copilot Chat session 内部复盘,不用于项目阶段复盘。项目阶段复盘请使用 retrospective skill。** 输入:当前 Copilot Chat session(自动获取 transcript)。 输出:ai-gen-data/session-retro/{timestamp}/ 目录下的 report.md。
lzx-Bill/Project-Factory-Core/.agents/skills/express-incubate Skill
0★ repoUse when the user explicitly wants a fast, one-pass project incubation, such as "快速孵化", "快速出计划", "express plan", "帮我规划一下这个项目". This skill orchestrates the existing Project Factory skills with bounded multi-agent batches, dependency gates, document reviews, execution logging, a final project audit, and one implementation-kit handoff. Do not trigger it for ordinary single-stage planning or when the user requests gradual confirmation at every stage.
lzx-Bill/Project-Factory-Core/.agents/skills/implementation-kit Skill
0★ repoUse when generating an AI-ready implementation package from Project Factory design docs — "generate implementation kit", "create handoff", "打包实现", "生成实现套件", "export to implementation", "交接". This skill reads ALL design documents from a project directory and compiles them into multi-tool context files (CLAUDE.md, AGENTS.md, Cursor rules, and Copilot instructions) plus per-task prompts and acceptance criteria. Output is a self-contained implementation-kit/ directory that can be copied to any new project directory and executed by any AI coding tool.
lzx-Bill/Project-Factory-Core/.agents/skills/plain-language-handoff Skill
0★ repoUse when generating a plain-language project summary that anyone can understand — "generate plain summary", "create readable handoff", "non-technical summary", "全员可读概要", "通俗交接文档". This skill compiles a comprehensive yet accessible document from project design docs, written in everyday language that serves both human stakeholders (no technical background needed) and AI agents (complete enough for full design and implementation). Output: implementation-kit/PLAIN-LANGUAGE-SUMMARY.md
lzx-Bill/Project-Factory-Core/.agents/skills/prompt-authoring Skill
0★ repoUse when writing structured implementation prompts, packaging context for AI coding, mapping prompts to tasks, or refining prompt granularity. Trigger when user says "write a prompt", "implementation prompt", "handover prompt", "context packaging", "task prompt", "decompose into prompts", or when moving from delivery-planning to actual prompt generation. The output goes into prompts/ directory. Coordinate with data-api-design for API consistency and with acceptance-design for prompt-to-acceptance mapping.
lzx-Bill/Project-Factory-Core/.agents/skills/skill-creator Skill
0★ repoCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit or optimize an existing skill, run evals to test a skill, or optimize a skill description for better triggering accuracy. Also use when a gap in the current skill set is identified.
kody-w/rapp-skills/film-m365 Skill
0★ repoFilm a Copilot Studio agent in the REAL M365 Copilot surface — the ONLY sanctioned way to make agent demo videos. Playwright drives a dedicated authed Chrome profile, records the page via CDP (no screen capture, no desktop takeover), then trims, dead-time-compresses, and gate-verifies to GO. Use when Kody says "film <agent>", "make a video of <use case>", "location shoot", or any ask for an agent demo video. NEVER use the synthetic stage — qa_gate blocks it (synthetic-stage check).
kody-w/rapp-skills/mcs-deploy Skill
0★ repoTake a built Copilot Studio solution (or build one from agent.py files first), import it with a FRESH version name, publish in the correct order, verify it actually answers, and hand back the maker-portal link. Never asks Kody for creds that live in local.settings.json. USE THIS SKILL when Kody says: "deploy it to copilot studio", "deploy the vN so I can test", "import the solution", "publish the bots", "test it over direct line", "make sure this copilot solution actually does what it says", or asks for a fresh install of a solution.
kody-w/rapp-skills/muscle Skill
0★ repoRun the CSM delegation loop: hand grunt/bulk work (building, drafting, coding, doc generation, deep debugging) to GitHub Copilot CLI GPT-5.6 Sol as a background muscle job, then gate-check the FULL result hands-on with real inputs, fix via --resume rounds until it stands, and add the Fable-only layer last. USE THIS SKILL when Kody says: 'delegate to sol', 'muscle this', 'have sol/copilot build it', 'spawn the muscle', 'send it to the muscle', 'copilot the grunt work', 'ration fable', or whenever a task is bulk artifact generation (code, tests, docs, ports, sweeps) that does not need Fable-level judgment to produce — the output is an artifact, not a decision. Also use when a hands-on verification loop exceeds ~2 probes on the same failure (deep debugging is muscle work).
kody-w/rapp-skills/rapp-pipeline Skill
0★ repoRun RAPP (Rapid Agent Prototyping Platform) pipeline to generate AI agents from transcripts. Use when building agents, generating Copilot Studio solutions, or deploying to Microsoft AI stack.
lestermarch/postgres-ai-experts/skills/azpg-pgvector-rag Skill
no license0★ repoDesign and implement vector / RAG search on Azure Database for PostgreSQL Flexible Server using pgvector, the azure_ai extension (in-database Azure OpenAI embeddings) and DiskANN / HNSW / IVFFlat indexing. Use this skill whenever the task involves semantic search, embeddings, RAG retrieval, similarity search, "nearest neighbour", a vector column, an embeddings pipeline, or vector index tuning on Azure Postgres — even when the user does not say the word "pgvector". Covers extension allow-listing, CREATE EXTENSION, azure_ai.set_setting, azure_openai.create_embeddings, choosing and building a vector index, and writing the retrieval query.
Paldom/promptimize/skills/engineer-agent-loop Skill
0★ repoDesigns and diagnoses the control flow of agentic loops - pattern selection, supervisor-enforced caps and terminal states, no-progress detection, external verifier gates, context and compaction budgets, crash recovery. Use when an agent loop stalls, drifts, repeats itself, or never stops, or when building agent orchestration. Not for goal or prompt wording, ordinary code loops, or scheduled jobs.
Paldom/promptimize/skills/evaluate-prompts Skill
0★ repoBuilds standalone prompt evaluations - test sets from real failures, paired A/B comparisons with sample-size math, deterministic checks before LLM judges, judge calibration and bias controls. Use when asked which prompt is better, to A/B test or benchmark prompts, or to set up LLM-as-judge scoring. Not for rewriting prompt text, agent goals, or loop runtime design.
Paldom/promptimize/skills/optimize-prompt Skill
0★ repoRewrites and tunes existing prompt text - removes contradictions, applies reasoning-vs-standard regime rules, sizes few-shot examples, fixes structured output, follows current Fable 5 / GPT-5.6 guidance, then checks the rewrite. Use when asked to optimize, improve, rewrite, shorten, tune, fix, or migrate a prompt or system prompt. Not for agent-run goal text, prompt benchmarks, or loop design.
Paldom/promptimize/skills/write-agent-goal Skill
0★ repoDrafts or fixes the goal text for an autonomous agent run (/goal, unattended sessions) - checkable success conditions, evidence requirements, scope and permission limits, retry and budget exits. Use when writing an agent goal, run definition of done, or run success criteria, or when a goal gets gamed or marked done early. Not for generic project goals, prompt rewrites, or loop runtime design.
Immersive-commons/ic-skills/skills/ic-onboarding Skill
0★ repoSet an agent's human user up on Immersive Commons end-to-end, explaining every step to them in plain English — mint a token, install + verify the MCP server, set their profile, EDUCATE them on everything IC offers, and file a quality tier-upgrade request. Use when the human says "set me up on IC", "connect me to immersive commons", "register my agent", "onboard me", "finish my IC setup", or hands you the verify URL. RFC 8628 device-code; two-phase because tiers are operator-approved. Official immersivecommons.com skill.
Immersive-commons/ic-skills/skills/ic-signed-agent Skill
0★ repoUpgrade an Immersive Commons agent token from bearer-only to bearer + Ed25519 signature (RFC 9421 strict subset). A leaked bearer alone is useless against tokens with signature enforcement on — the agent must also possess the private key. Use when the human says "make my IC agent more secure", "rotate my agent key", "sign my IC requests", or whenever the threat model includes token leakage (CI environments, multi-user dev boxes, shared agent runtimes). Requires the human to bind the public key via a Clerk-authenticated browser flow. Official immersivecommons.com skill.
Immersive-commons/ic-skills/skills/immersivecommons Skill
0★ repoStart here for immersivecommons.com — the official umbrella skill for Immersive Commons, the members-run AI builder space on Floor 10 of Frontier Tower SF. Routes an agent to the right IC skill (events, onboarding, headsets, rooms, inbox, leaderboard, feedback, THE SIGNAL, Z.ai keys, signed-agent hardening, operator admin) and explains the shared plumbing every one of them uses — the MCP server, agent tokens, tiers, and scopes. Use when the human says "set my agent up on Immersive Commons", "what can agents do at IC", "install the IC skills", or names immersivecommons.com without a more specific task.
Immersive-commons/ic-skills/skills/zai-keys Skill
0★ repoRequest and use a Z.ai (GLM) Claude-Code-compatible key from Immersive Commons — a 5-hour walk-in WORKSHOP pass (public tier, tied to an event) or a weekly-token MEMBER key (ic-member, pick a multiplier). Use when the human says "get me a Z.ai key", "I need a GLM key for the workshop", "set up Claude Code on the IC gateway", "request a zai key", "I'm at the IC event and want to code", or "give me a weekly Z.ai allowance". Walks the full path — pick key type → request via MCP → operator approves → configure Claude Code (4 ANTHROPIC_* env vars) → run, plus the time-box / weekly limits and what happens at the cap. Requires an IC agent token with `keys:request` (granted at every tier). Official immersivecommons.com skill.
cekrauseee/harness/skills/harness-worktree Skill
0★ repoPlan, create, validate, adopt, and retire isolated Git worktrees with Harness-controlled semantics and lifecycle while delegating only physical path and storage to the active agent host. Use when naming, creating, locating, validating, adopting, reusing, or removing a worktree, or when a task needs the Harness `type/slug` branch convention.
proluct/skill-press/skills/skill-press Skill
0★ repoThe Skill Printing Press. A guided factory that turns an idea into a ship-ready Claude Code skill, with prior-art research, one plan-approval gate, and subagent-tested verification before anything counts as done. Use when the user wants to create, build, or make a new skill, says "turn this into a skill", "skill for X", "capture this workflow", or types /skill-press. Not for editing an existing skill (use skill-press-polish), not for hooks or settings (those live in settings.json), not for subagent definitions.
mrcha033/skills/plugins/learnus-course-copilot/skills/learnus-course-copilot Skill
no license0★ repoUse the user's authenticated browser session to inspect Yonsei LearnUs courses, collect authorized course materials, identify assignments and deadlines, and produce study-ready structured output. Use for LearnUs course pages, lecture resources, VOD links, assignment tracking, or downloads when credential-based scripts fail or SSO state must remain in the browser.
mrcha033/skills/skills/agent-finish-line Skill
no license0★ repoFinish an implementation, experiment, migration, release, or debugging task with the least process needed. Use when work is expanding, repeating unchanged attempts, or stopping before the requested deliverable; drive it to completion or one evidence-backed external blocker without creating task-management artifacts.
mrcha033/skills/skills/learnus-course-copilot Skill
no license0★ repoUse the user's authenticated browser session to inspect Yonsei LearnUs courses, collect authorized course materials, identify assignments and deadlines, and produce study-ready structured output. Use for LearnUs course pages, lecture resources, VOD links, assignment tracking, or downloads when credential-based scripts fail or SSO state must remain in the browser.
arkaigrowth/agent-skills/mac-sys-diagnostics Skill
0★ repomacOS system diagnostics for memory leaks, zombie processes, MCP server issues, WindowServer problems, and Secure Input keyboard issues (Caps Lock failures). Use when user reports system slowness, high memory usage, zombie processes, keyboard malfunctions, or requests system health checks. Provides copy-paste command blocks with clear sections for progressive diagnosis and remediation.
atomgunlk/skills/skills/engineering/use-worktree Skill
0★ repoUse when working with Git worktrees — triggers include "create worktree", "add worktree", "sibling worktree", or a plan file path. ALWAYS use for managing Git worktrees.
TJCurnutte/cohort-method Skill
0★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".
avnath13/evalpilot/skills/eval-audit Skill
0★ repoDiagnose an EXISTING eval pipeline and produce findings ranked by business impact. Use when the user inherited or already has evals (traces, judges, dashboards, labeled data) and wants to know what's wrong or what to fix first. Do NOT use to build evals from scratch (use prepare-data / user-simulator) or to grade (use grade), this only audits.
avnath13/evalpilot/skills/grade Skill
0★ repoGrade agent traces against rubrics using deterministic checks and binary LLM-as-judge (critique-then-label), no external eval framework required. Use when the user wants to score, judge, or evaluate agent outputs against criteria. Do NOT use to author a judge or validate it against human labels (use validate-evaluator/calibrate), and do NOT use to fix failures (use analyze-failures then optimize). Prefer a deterministic check over a judge whenever the criterion is objective (fix-before-eval).
aoli0919/robotics-project-starter-skills/skills/webcam-agent-lab Skill
0★ repoPlan a webcam-based perception-action toy project without robotics hardware.
pecigonzalo/agent-skills/skills/role-orchestrator Skill
0★ repoUse this skill when a primary agent coordinates subagents. Provides delegation criteria, quality gates, TODO-Store linking, and communication patterns.
rvdsmaile/working-backwards/.agents/skills/handoff-working-backwards Skill
no license0★ repoTurn an approved Working Backwards initiative into a product brief and build plan. Use after announcement, FAQ, decisions, and critical critique findings are complete; do not use to write implementation code or estimates.
Marineconnect/marineconnect-agent-skills/plugins/marineconnect-ops/skills/marineconnect-mcp-safety Skill
no license0★ repoApply MarineConnect MCP safety rules. Use for production MarineConnect operations, MCP QC, OAuth-connected MarineConnect sessions, SuperAdmin permission checks, read-safe Phase 2 workflows, and any request that could accidentally trigger write, reboot, cancel, opt-in, opt-out, deactivate, delete, or pricing mutation behavior.
aoli0919/prompt-debugging-skills/skills/prompt-test-case Skill
0★ repoCreate test cases, expected behaviors, and failure probes for a prompt.