SKILL.md packages that extend Claude Code, Cursor, Copilot, and other AI agents.
Tags
skillshub
Guided workflow for drafting high-quality technical articles for developer audiences: intake, title/hooks, structure, drafting, images, and finalization.
knowledge-work-plugins
Reads forwarded customer emails, looks up PayPal and HubSpot records, drafts a tone-matched reply in the owner's voice, and can stage or issue refunds with expl
vexjoy-agent
Plans multi-part content series: assesses viability, selects structure and cadence, and generates detailed per-part plans with cross-linking and publishing sche
workflow-architect
Lightweight 2-phase software architect workflow (Understand → Execute) for fast project planning and execution with explicit checks and minimal state.
dgk-gpt
Temporarily bias the agent toward a research-heavy workflow: plan sub-questions, retrieve sources, synthesize with explicit citations, and verify before impleme
ospec
CLI-driven, document-first workflow for initializing and maintaining change-ready projects and project knowledge with OSpec.
awesome-omni-skills
Git-aware revert workflow for reverting tracks, phases, or tasks while preserving provenance and upstream workflow; includes safety checks and an execution plan
awesome-omni-skills
Workflow and best practices for restoring and preserving project context during code refactoring, with provenance and selective file restoration.
montaj
Agent-first video editing toolkit: probe, trim, transcribe, caption and materialize clips via CLI or HTTP workflows for automated editing pipelines.
claude-code-plugins-plus-skills
Image editing and image-to-image workflows for Ideogram: inpainting (Magic Fill), style remix, upscaling, image description, and reframing for new aspect ratios
swarms
Facilitator-focused procedure to close stuck review findings by requiring concrete evidence or concession, enforcing finality once decided.
twinmind
Project lifecycle manager for TwinMind: create, pause, resume, complete, archive projects; manage actions and tasks and trigger post-op pipelines.
multi-agent-shogun
Design, validate and review Claude Code-style SKILL.md files: frontmatter guidance, trigger design, testing patterns, and recommended file structure for robust
tackline
Run multi-step research and analysis chains (gather, distill, rank, verify, plan) automatically; choose and orchestrate primitives based on a user goal.
awesome-omni-skills
A comprehensive penetration testing checklist and operational workflow to plan, scope, monitor, and remediate security tests while preserving provenance of upst
vvm
A small domain-specific language and runtime for writing agentic programs (.vvm) where the LLM is the execution runtime; provides compile, run, and inspection c
superpowers
Automate Skyfire operations through Composio's Skyfire toolkit using Rube MCP; discover tools, verify connections, and execute schema-compliant workflows.
databricks-lab
Adopt the Agents Orchestrator role: follow the repository's agent definition to plan, coordinate, and produce orchestrated agent deliverables.
awesome-omni-skills
Selects and places the right type of social proof (testimonials, logos, case studies) to close specific trust gaps while preserving provenance and upstream work
bkit-gemini
Run prompts or shell commands on a recurring interval (seconds/minutes/hours) with safety limits and optional stop conditions — useful for monitoring and period
claude-build-workflow
Convert product requirement documents (PRDs) or Markdown specs into Ralph's prd.json format, producing small verifiable user stories for autonomous execution.
skillshub
A collection of project initialization and workflow-generation skills that scaffold new projects, produce CLAUDE.md/AGENTS.md/Plans.md, and automate standard se
projectmnemosyne
Capture session learnings and turn them into reusable skill plugins (PR-ready) to preserve team knowledge and avoid repeating mistakes.
claude-skills-pack
Defines an inter-skill artifact protocol: how skills produce, publish, discover, and consume handoff manifests to enable reliable skill chaining and artifact li
agentic-dev-team
A four-phase debugging protocol (reproduce, investigate, root-cause, fix) that enforces evidence-driven fixes and prevents guess-and-fix thrashing.
anvil
Pick up a paused long-running task, make measurable progress, and save a concise breadcrumb file describing work done and what's left so another agent can conti
