Agents
Build and improve your own AI agents. 637 skills
#1agent-architecture-audit
Makes your agent check if everything is working correctly. It finds and fixes problems in AI systems, ensuring they perform as expected.
affaan-m/ECC
#2agent-eval
Helps your agent compare different AI coding tools. It measures which one is faster, cheaper, and more reliable for specific tasks.
affaan-m/ECC
#3agent-harness-construction
Makes your agent better understand how to perform tasks. It optimizes the tools and instructions the agent uses to increase its chances of success.
affaan-m/ECC
#4agent-introspection-debugging
Helps your agent fix itself when something goes wrong. It investigates failures to understand the cause and find a solution, rather than just trying again.
affaan-m/ECC
#5agent-security-hardening
Makes your agent more secure when handling sensitive information or external tools. It protects against unauthorized access and errors.
affaan-m/ECC
#6agent-self-evaluation
Helps your agent evaluate its own work. It checks if the answers are correct, complete, and clear, suggesting improvements for the future.
affaan-m/ECC
#7agentic-engineering
Makes your agent work smarter and more cost-effectively. It plans tasks, breaks them into smaller parts, and chooses the best AI tool for each step.
affaan-m/ECC
#8ai-first-engineering
Helps teams use AI to create code. It defines how AI agents work alongside humans, ensuring the generated code is high quality.
affaan-m/ECC
#9blueprint
Helps your agent create a detailed plan for large projects. It breaks down work into smaller parts and organizes execution to ensure success.
affaan-m/ECC
#10cost-aware-llm-pipeline
Makes your agent use AI models more cheaply. It chooses the right model for each task and controls spending.
affaan-m/ECC
#11foundation-models-on-device
Lets your agent use language models directly on Apple devices. It generates text, uses tools, and streams data in real-time within iOS apps.
affaan-m/ECC
#12mcp-server-patterns
Lets your agent build MCP servers using Node/TypeScript. Helps set up tools, resources, and data validation for the server.
affaan-m/ECC
#13ml-adoption-playbook
Helps your agent add artificial intelligence to existing programs. Guides from the initial idea to integrating a first model.
affaan-m/ECC
#14mle-workflow
Enables your agent to manage the entire machine learning lifecycle in production, from data handling to monitoring and rollbacks.
affaan-m/ECC
#15regex-vs-llm-structured-text
Lets your agent choose the best way to read texts like forms or notes. It uses simple rules or AI to extract information.
affaan-m/ECC
#16agent-payment-x402
Allows AI agents to make payments with per-task budgets and spending controls. It uses non-custodial wallets and supports various blockchain networks for payments.
affaan-m/ECC
#17agent-runtime-gateway-smoke-test
Verifies a local agent's API, a temporary gateway tunnel, and a remote callback. It runs a simple task to ensure everything is working correctly without extra tools.
affaan-m/ECC
#18agent-sort
Helps create an installation plan for a repository by organizing skills and commands. It separates daily use items from library items to optimize usage.
affaan-m/ECC
#19agentic-os
Enables building persistent multi-agent operating systems using Claude Code. It manages memory, automation, and state without external databases, ideal for complex systems.
affaan-m/ECC
#20ai-regression-testing
Helps your agent perform regression testing on AI-generated code, avoiding blind spots.
affaan-m/ECC
#21autonomous-agent-harness
Transforms Claude Code into an autonomous agent with memory and scheduling. It allows the agent to operate continuously, run scheduled tasks, and self-manage.
affaan-m/ECC
#22autonomous-loops
Helps your agent create complex, autonomous workflows. It lets the agent perform tasks in steps, independently, to achieve a goal.
affaan-m/ECC
#23continuous-agent-loop
Helps your agent self-check and recover from failures. It's useful for ensuring the agent runs reliably and continuously.
affaan-m/ECC
#24continuous-learning-v2
Lets the agent learn from interactions and create new skills. It's useful for improving the agent's performance over time.
affaan-m/ECC
#25continuous-learning
This is an older version for agent learning. Use the newer version for learning from sessions and extracting patterns.
affaan-m/ECC
#26council-multi-model
Adds a critique from another model for important decisions. It helps ensure safer and more well-thought-out decisions.
affaan-m/ECC
#27council
Gathers a group of models to make difficult decisions. It's useful when there are multiple options and structured analysis is needed.
affaan-m/ECC
#28counterparty-channel-discipline
Controls the agent's communication in shared chats. It ensures the agent speaks only when needed and doesn't leak information.
affaan-m/ECC
#29eval-harness
Lets your agent use an evaluation system for development. It defines success criteria and measures AI task reliability before and after coding.
affaan-m/ECC
#30exa-search
Lets your agent search the web, code, and company information using AI. It's useful for quickly finding relevant data and examples for your tasks.
affaan-m/ECC
#31fal-ai-media
Helps your agent create images, videos, and audio using AI. It allows generating multimedia content from text descriptions or other media.
affaan-m/ECC
#32gan-style-harness
Allows your agent to build high-quality applications autonomously. It uses a generation and evaluation system to ensure the output meets a quality standard.
affaan-m/ECC
#33gateguard
Prevents the agent from performing write or code actions without first providing concrete information. It ensures the agent has the necessary data before proceeding, improving quality.
affaan-m/ECC
#34iterative-retrieval
Helps your agent search for information more intelligently. It refines the search in multiple steps to find what it needs.
affaan-m/ECC
#35llm-trading-agent-security
Lets your agent handle trading security. It protects against fraud, sets limits, and verifies transactions for autonomous trading agents.
affaan-m/ECC
#36loop-design-check
Lets your agent design or check operational loops. It ensures agents achieve goals without failures or unwanted behaviors.
affaan-m/ECC
#37recsys-pipeline-architect
Helps your agent design systems that recommend items. It uses a framework to select the best content, like in social media feeds or search systems.
affaan-m/ECC
#38recursive-decision-ledger
Has your agent perform repeated tests safely. It records all decisions and checks, ensuring important actions are only approved after much confirmation.
affaan-m/ECC
#39mcp-builder
Helps your agent create servers that connect LLMs to external services. Essential for integrating APIs and services using Python or Node/TypeScript.
anthropics/skills
#40prepare-mcp-integration
Prepares vendor (MCP) integrations through research and human approval. It follows a guided process for new integrations.
paperclipai/paperclip
#41first-task
Guides the agent on the user's first task. It interprets the initial answer, clarifies the goal, and proposes a plan or single task for approval.
paperclipai/paperclip
#42paperclip
Lets your agent interact with the Paperclip system to coordinate tasks. It's useful for checking assignments, updating status, commenting, delegating, or managing routines.
paperclipai/paperclip
#43agentmail
Allows your agent to read and send task emails. It can reply to emails and check if they were delivered, using its assigned inbox.
paperclipai/paperclip
#44complain
Lets your agent complain about tools or work. It sends a brief, direct complaint about issues, without permission, and continues the main task.
paperclipai/paperclip
#45paperclip-board
Helps your agent manage a company in Paperclip as a board member. It's used for onboarding, agent management, approvals, and work review.
paperclipai/paperclip
#46paperclip-converting-plans-to-tasks
Lets your agent decide if Paperclip work needs delegation. It connects justified tasks with owners and dependencies when turning a plan into work.
paperclipai/paperclip
#47paperclip-create-agent
Helps your agent create new agents in Paperclip. It inspects configurations, compares options, and submits a request to hire a new agent.
paperclipai/paperclip
#48html-ppt-hermes-cyber-terminal
Lets your agent explain how to choose and connect your own AI models. It covers costs, quality, and routing decisions, useful for engineering and IT teams interested in AI.
nexu-io/open-design