Skills
96,981 packaged capabilities and playbooks for agents. Install any of them with one command.
Agent workflows
This skill should be used when the user asks to "create a PRD for Replit", "write product requirements", "document app specifications", "prepare a product spec", or needs a comprehensive specification document that Replit Agent can use for complex, multi-feature applications. Creates structured PRDs optimized for AI-powered development.
Proactively analyze user requests at the start of conversations to determine task type, assess prompt quality, and intelligently recommend which skills to activate. Should activate for ALL user requests to ensure optimal workflow. Evaluates clarity, specificity, and completeness to suggest prompt-optimizer when needed. Identifies UI design tasks for ui-analyzer and component requests for react-component-generator. Acts as intelligent skill coordinator.
Selecting and dispatching the right Ring developer specialist agent (backend Go/TS, frontend, UI designer/engineer, Helm, frontend QA, prompt reviewer) for a technology task. Use when you need deep stack expertise and must decide which agent to invoke via the Task tool, including parallel dispatch of independent work. Skip for general code review (use ring:reviewing-code).
When a user is stuck, frustrated, or describing a problem vaguely, do NOT immediately suggest solutions. First, force structured problem articulation through targeted questions: What did you expect? What happened instead? What have you tried? Only after the problem is clearly defined, propose solutions.
Build and maintain Ruby or Rails applications with the RubyLLM AI framework. Use for chats, agents, tools, MCP servers, structured output, judgments, evaluations, media generation, transcription, OCR, moderation, embeddings, reranking, tracing, Rails integration, and RubyLLM upgrades; not for contributing to the framework itself.
Capture user feedback and corrections to enhance repository documentation and skills. Transforms conversational feedback into persistent knowledge by updating the appropriate CLAUDE.md files, skills, or other documentation. Use when: (1) User provides a correction or clarification, (2) User says "remember this", (3) User provides feedback about how something should work, (4) After completing work where new patterns or knowledge were discovered, (5) User explicitly asks to update documentation. Triggers: "remember this", "update the skill", "add this to documentation", "you should know", "in the future", "always do", "never do", "that's wrong", "actually it should be", "/self-improvement", "capture this", "document this pattern", "add to CLAUDE.md"
Captures learnings, errors, and corrections to enable continuous improvement. Use when: (1) A command or operation fails unexpectedly, (2) User corrects Claude ('No, that's wrong...', 'Actually...'), (3) User requests a capability that doesn't exist, (4) An external API or tool fails, (5) Claude realizes its knowledge is outdated or incorrect, (6) A better approach is discovered for a recurring task. Also review learnings before major tasks.
Structured reflective problem-solving methodology. Process: decompose, analyze, hypothesize, verify, revise. Capabilities: complex problem decomposition, adaptive planning, course correction, hypothesis verification, multi-step analysis. Actions: decompose, analyze, plan, revise, verify solutions step-by-step. Keywords: sequential thinking, problem decomposition, multi-step analysis, hypothesis verification, adaptive planning, course correction, reflective thinking, step-by-step, thought sequence, dynamic adjustment, unclear scope, complex problem, structured analysis. Use when: decomposing complex problems, planning with revision capability, analyzing unclear scope, verifying hypotheses, needing course correction, solving multi-step problems.
Discover and list recent AI coding sessions across Claude Code, Claude Desktop, and OpenCode on Windows, macOS, and Linux. Scripts extract project paths, session IDs, and timestamps from session artifacts. PowerShell for Windows, Bash for macOS/Linux. Use when: finding recent coding sessions, reviewing AI session history, exporting session logs to CSV, troubleshooting missing sessions, or auditing Claude Code/OpenCode usage across environments.
Register and configure an AI agent on OpenAnt. Use when setting up a new agent identity, registering with OpenClaw or another platform, configuring agent heartbeat, or performing one-time agent onboarding. Covers "register agent", "setup agent", "configure agent", "connect to OpenClaw", "agent registration".
Configure this repo for the Gharib89/skills engineering skills: draft its ship profile, local gate, PR template, coding-standards doc and reviewer scaffolding, and check the host tooling. Run once after /setup-matt-pocock-skills, before the first /ship.
Judge subskill for simmer. Scores a candidate artifact against user-defined criteria on a 1–10 scale and produces ASI (highest-leverage direction) for the next generator round. Supports judge-only, runnable evaluator, and hybrid evaluation modes. Do not invoke directly—dispatched as a subagent by the simmer orchestrator.
Build persistent LLM agent memory with SimpleMem - semantic lossless compression with 30x token reduction. MCP server (cloud/self-hosted), Python API, cross-session memory. Use when: adding long-term memory to agents, cross-session context recall, semantic dialogue search, memory consolidation.
Create new Skills for Claude Code including YAML frontmatter, descriptions, instructions, and structure. Use when creating, building, or designing skills, or when asked about skill creation, structure, best practices, or debugging skills that don't activate properly.
Run the AI team pipeline for squad tasks — orchestration loop with 2 agents (Worker, Reviewer), single-step execution, and code review. Use /squad-run to execute tasks through the 7-column pipeline. AUTO-TRIGGER when: user says "implement task NNN" or any task ID + implement/build/do combination; or user confirms with "yes/ok/go/do it" after Claude proposes implementing a specific squad task.
(forwward) Onboards founders to forwward-teams by learning what they build, initializing the environment, and recommending which skills to use first. Triggers on first install, new project setup, getting started, or when a user asks what forwward-teams can do.
Pause and critically reflect on the direction of the current conversation. Use this whenever the user says 'take a step back,' 'let's zoom out,' 'are we sure about this,' 'wait — am I missing something,' 'sanity check this,' 'pressure test this,' 'are we going down the wrong path,' or expresses doubt about where the discussion has gone. Also use proactively if you notice the conversation has accumulated assumptions that haven't been examined, or has drifted from the user's original goal. Output is a structured reflection — original goal, path taken, assumptions made, what might be wrong, and a recommendation.
Submit completed work for a task on OpenAnt. Submission = text description + files. IMPORTANT — before submitting, always check if your work produced any files and upload them first. Use when the agent has finished work and wants to deliver results, submit a solution, turn in deliverables, upload files, or send proof of completion. Covers "submit work/task", "deliver results", "I'm done", "here's my work", "submit solution", "upload and submit", "attach proof", "deliver file", "send deliverable".
REQUIRED for every code-related task. Read SKILL.md before acting, then start the reply with workflow triage. Use `Workflow mode: Lightweight — <reason>; success: <outcome>; plan: no.` for clear localized changes and read-only review or discovery. Use the Full triage block and persisted plan for risky, coupled, cross-session, or contract-changing implementation. Escalate from Lightweight before target-file edits when discovery reveals Full-mode risk. Focused delegated agents follow Delegated Mode instead.
(forwward) Exports all stored memories, context, preferences, and instructions for migration to another AI platform. Triggers on platform migration, backing up AI context, auditing stored data, or setting up a new AI assistant with existing preferences.
Discovers your full tool environment and amplifies prompts with capability awareness. Suggests optimal tool compositions as non-binding options. Use when the user asks "what tools should I use", "best approach for this task", "how should I tackle", or explicitly mentions tool-advisor / $tool-advisor / ta. Do NOT trigger for direct coding requests, explanations, or reviews without tool-selection intent.
Use when the user wants an autonomous ML research loop that pressure-tests competing ideas before spending compute — several research subagents each propose one architecture change, a self-calibrating Judge critiques them against a rubric, the proposers refine, and the Judge picks the single change to run. The Judge learns to pick better over time by scoring its own predictions against realized metric deltas, recording predicted-vs-realized in a calibration ledger and refining its working rubric. The result is an experiment ledger where each iteration's change won a de-biased tournament. Not for running a single pre-decided experiment, and not for analysis-only exploration — for one hypothesis proposed and run per iteration without competition, use the sibling ml-autoresearch loop.
A live, growing registry of agent skills published by Anthropic, OpenAI, Vercel, Microsoft, Cloudflare, Stripe, HashiCorp, Supabase, and many others. Search it before starting any non-trivial task, fetch the top result, follow it instead of going freehand. After using a skill, report back so the registry ranks it correctly for the next agent. If you build something reusable, ask the user, then upload it.
Deliver the verdict for a unit of work once the TESTER and DEV have both submitted: run the TESTER's tests against the DEV's code, arbitrate each failure against the contract (the code deviated, the test deviated, or the contract did not specify it), verify that each role touched only its own paths, and produce the traceability matrix for the PR. Use when the LEAD has received both deliveries (ENTREGA TESTER and ENTREGA DEV) for a unit of work, or when asked to verify, judge, arbitrate, or give the verdict on a team delivery
Apply James's AI trust principle to both outputs and autonomous actions. Use when generating or executing consequential changes, delegating tasks or assessing reliability; require independent verification, bounded permissions/time/data/money and recoverability proportionate to impact.
A browser-based web access tool. Use this skill when a real browser environment is needed, including: operations requiring a logged-in session, platforms with strict anti-scraping measures (Xiaohongshu, WeChat official accounts, etc.), page interaction, and scraping dynamically rendered content. For simple searches and public page scraping, the built-in WebSearch/WebFetch can be used directly without loading this skill.
Live widgets for the Iskron window instead of a recap — only where the answer renders an Iskron window: home agent, graph window. Triggers: “show graph cases,” “which cases are open,” “cases for this node,” “case card #N,” “where is the case now,” “which agents are active,” “who is working in the graph now,” “what are the agents doing.” Places one of the cases, case, agents widgets as a block with the iskron label and realm parameter. Don’t load in Claude Code, Telegram, or other places where the block will appear as code.
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.