One search for your own team's files and the public ecosystem. See how real projects write their agent instructions, and bring the good ideas home.
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
jovd83Skilljovd83/junit5-skill/SKILL.md
Use when the agent needs to create, run, evaluate, debug, correct, modernize, document, or route JUnit 5 tests and related workflows across component, integration, slice, repository, service, regression, smoke, and migration scenarios.
jovd83Skilljovd83/lifelike-synthetic-data-generator/SKILL.md
Generate realistic, population-shaped synthetic datasets for tests, demos, and fixtures. Trigger for CSV/JSON exports, locale-aware personal/company fields, seeded generation, custom regex formats, Belgian identifiers (INSZ, eID), and distribution-backed sampling. Runs a local generator instead of fabricating in-context.
jovd83Skilljovd83/modern-dependency-guard/SKILL.md
Review, validate, and modernize technology choices before suggesting code that depends on external packages, frameworks, SDKs, CLIs, or hosted services. Use to choose a stack, sanity-check dependencies, replace deprecated tools, suggest modern alternatives, or scan a Node.js package manifest.
jovd83Skilljovd83/new-feature-sdlc-skill/SKILL.md
Orchestrate approved end-to-end new-feature or change-request delivery in an existing codebase: discovery, planning, implementation, testing, closeout. Skip for bug fixes, refactors, test-only work, ops incidents, or unapproved features.
jovd83Skilljovd83/plantuml-skill/SKILL.md
Use when a diagram must be written and rendered in PlantUML: sequence, C4 architecture (context, container, component), state, class or activity diagrams from notes, specs or code. Writes the .puml source with the house theme and renders SVG and PNG locally, setting up Java and Graphviz on first run. Also for requests like render this .puml, draw a C4 container diagram, or fix PlantUML syntax errors. To choose between Mermaid, PlantUML and draw.io first, use diagram-generator.
jovd83Skilljovd83/principal-audit-refactor/SKILL.md
Audit a local software project, produce a severity-ranked engineering review, and execute an approval-gated refactor plan. Use when the agent needs to assess prototype-quality or inconsistent codebases, run stack-aware checks, write timestamped audit artifacts, and then implement production-grade improvements with explicit safety boundaries, report contracts, and refactoring discipline.
jovd83Skilljovd83/prompt-efficacy-scorer/SKILL.md
Audit and improve prompts for clarity, consistency, instruction hierarchy, verbosity, and model fit. Use to review, score, tighten, debug, or optimize a prompt, system prompt, agent instructions, or template, or deliver a scorecard plus an optimized rewrite.
jovd83Skilljovd83/retro-board-creator/SKILL.md
Use when creating fun, visually engaging agile or scrum retrospective boards in Mermaid, Draw.io, PNG/SVG, Microsoft Whiteboard guidance, or similar formats, and when interpreting completed retro boards from photos, screenshots, Mermaid, Draw.io, or text to produce summaries, action items, follow-up notes, Jira/Confluence-ready output, or facilitation artifacts.
jovd83Skilljovd83/skill-yaml-cleanup/SKILL.md
Lint AgentSkill folders before they are installed, synced or published: frontmatter that parses as YAML, a name that matches the folder, a description within 1,024 characters, no development-tree paths, no model-run telemetry instructions, and optionally the README standard. Also trims oversized frontmatter, with approval.
jovd83Skilljovd83/slide-deck-preparation/SKILL.md
Convert source material into grounded, presentation-ready slide deck outlines. Use when Codex needs to plan, audit, restructure, or enrich a presentation from URLs, articles, papers, transcripts, notes, documents, or an existing deck outline, with explicit audience framing, narrative arc, slide archetypes, speaker notes, visual guidance, and source citations.
jovd83Skilljovd83/tdd-test-case-reviewer/SKILL.md
Use when the user wants to review, audit, critique, score, or mentor drafted software test cases, manual scenarios, UAT cases, or test suites for requirement traceability, coverage gaps, technical correctness, standards compliance, weak expected results, or TSS or TDD-style test design quality before execution, automation, or sign-off.
jovd83Skilljovd83/test-analysis-skill/SKILL.md
Use when reviewing use cases, user stories, acceptance criteria, business requirements, or specifications for testability, ambiguity, requirement quality, delivery risk, or missing decisions before implementation or test design. For static reviews, risk-based analysis, and readiness triage. Do not use when writing tests or code.
jovd83Skilljovd83/test-artifact-export-skill/SKILL.md
Keep test cases and results in sync with test-management tools, and render approved test cases for review. Three jobs - export approved cases as detailed or TDD markdown, summary tables, plain text, BDD/Gherkin, Xray .feature bundles, Zephyr Scale CSV, or TestLink and TestRail mappings; map the IDs a tool assigned (TestRail, Xray, Zephyr Scale, TestLink) back into local tests and docs; and publish Playwright, Cypress, Rest Assured or JUnit results into those tools. Use when the test logic already exists. Not for designing tests or choosing techniques.
jovd83Skilljovd83/token-usage-cost-report/SKILL.md
Use when an agent must produce an auditable token-usage and cost report from structured runtime evidence, normalized usage bundles, or repository-level report sets. This skill discovers trustworthy evidence, resolves official pricing, refuses transcript-only estimation, and generates provenance-backed Markdown, HTML, and JSON artifacts with explicit exact, partial, estimated, or unsupported cost labels.
jovd83Skilljovd83/travelguide-copywriting-skill/SKILL.md
Use when writing high-quality travel guides, roadbooks, or itineraries. Sourced inputs can include coordinates, links, or place names. It supports personalization based on traveler ages, interests, and dietary needs, and uses live-verified TripAdvisor or Google reviews for restaurant recommendations.
Plain text files in a repository that tell a coding agent how the project works: commands to run, conventions to follow and things to avoid. CLAUDE.md, AGENTS.md, cursor rules and skills are the common kinds.
CLAUDE.md is read by Claude Code. AGENTS.md is an open format that Codex, Cursor and other agents read. Many projects keep one and point the other at it.
A folder with a SKILL.md that describes one capability, such as filling PDFs or reviewing code. The agent loads it only when the task calls for it.
Your agents already can, over MCP, limited to the files you're allowed to read. Searching them from this page is coming.
