Tag: llm-agent
Agent lifecycle hooks for observation and interception
Provides comprehensive hooks to observe and intercept the entire agent execution lifecycle, covering steps, tool calls, context compression, and human intervention. This mechanism allows developers to build advanced tracing, mocking, and cu…
Deep Automated CSV and Excel Data Analysis Tool
This tool provides deep, automated data exploration for CSV, Excel, and TSV files, generating comprehensive, interactive web reports. It extracts statistical features, identifies anomalies, and provides actionable business insights across d…
Guide to building production MCP servers
This guide outlines mandatory best practices and architectural patterns for developing robust MCP servers using the mcp-use framework. It provides detailed reference material for implementing tools, data resources, prompts, and interactive …
Reproduce Research Methods and Benchmark Metrics
This skill automates the process of implementing a new research method into a structured Jupyter notebook, ensuring reproducibility by utilizing existing data splits and dependencies. It systematically records all required metrics and imple…
Zero-API-Key Web Search and Evidence Verification
This comprehensive CLI tool provides a structured workflow for source-backed research, enabling multi-engine web searches, deep page browsing with auto-unlocking, and evidence-aware claim verification. It generates citation-ready reports, s…
Interactive PR comment resolution workflow
This skill systematically addresses all unresolved pull request review comments by fetching thread details via GitHub GraphQL API. It guides the user through reviewing the issue, researching documentation, and applying local code fixes or r…
Agent Voice Call Management Tool
This tool enables agents to initiate, manage, and terminate voice calls using multiple providers such as Twilio, Telnyx, and Plivo. It provides structured actions for sending messages, continuing conversations, and checking the current call…
Agentic Memory Protocol for Context Management
This protocol guides agents on implementing robust, persistent memory context by defining when and how to search and write memories. It mandates deliberate context retrieval using structured metadata filters and ensures key learnings, decis…
Web3GPT Agent Interaction Skill
An API-based skill for initiating and managing conversational sessions with Web3GPT agents. It enables developers to send messages, track chat history, and trigger blockchain deployments via natural language.
Self-learning memory layer for coding agents
Instinct is a local memory layer that observes coding patterns, tracking confidence through repeated usage. It automatically promotes recurring fixes and preferences into actionable rules and suggestions for future sessions.
Pre-action gates for preventing agent mistakes
This skill implements pre-tool-use hooks to intercept and block known bad agent actions, such as repeated hallucinations or destructive calls. It saves tokens and prevents the model from executing mistakes that have been previously identifi…
Persistent memory and context management for LLMs
This skill provides persistent memory across sessions by mining completed interactions to extract durable lessons. It pre-loads relevant context at the start of new sessions, eliminating the need for repeated project explanations.
Zero-Trust Session Guardian for Agent Workflows
This skill executes a comprehensive, zero-trust audit at the start and end of every session, ensuring system integrity by cross-referencing environment variables, hardcoded URLs, and scheduler wiring against live API endpoints. It manages s…
Zero-API-Key Web Search and Evidence Verification
This tool facilitates a comprehensive, source-backed verification workflow, enabling multi-engine search and deep page reading. It supports evidence-aware claim checking and generates citation-ready reports, offering both free and professio…
Source-backed web search and evidence checking tool
This tool provides robust, source-backed web search and evidence verification without requiring API keys. It supports multi-engine SERP retrieval, deep page browsing with unlocker capabilities, and structured claim reporting for factual gro…
Adapt Ideas for Multi-Platform Content Creation
This skill transforms a single source asset, such as an article or podcast, into cohesive, platform-native content drafts. It ensures that outputs for channels like X, LinkedIn, and YouTube maintain distinct tones and structures, avoiding s…
orchestrates multi-step research pipelines
This skill orchestrates complex, multi-step research pipelines by executing deployed Mistral Workflows. It provides human-in-the-loop checkpoints, allowing developers to query progress, validate intermediate hypotheses, and inject sources b…
Dingo Article Fact-Checker
Extracts and verifies factual claims from markdown, text, or JSON files using web search evidence. The process generates a detailed report including accuracy scores and identified false claims.
Intelligent multi-account inbox review and prioritization
This skill performs a systematic, multi-account review of the inbox, grouping messages into threads and classifying them by priority (e.g., Action Required, FYI). It provides structured insights and suggests draft replies for immediate acti…
AI-powered professional networking agent for chat
Mingle integrates professional networking capabilities directly into the chat interface, allowing the AI to maintain a lightweight, privacy-preserving profile. It proactively surfaces relevant connections and facilitates double opt-in intro…
Local Agentic Knowledge Base Setup and Indexing
This utility provides a comprehensive, step-by-step wizard for setting up a local, agentic knowledge base. It handles backend initialization, document ingestion (including web crawling and PDF support), and integrates the resulting index wi…
Persistent GitHub PR Reviewer with Style Memory
This skill reviews GitHub Pull Requests, dynamically applying learned code style preferences and anti-patterns stored in persistent memory. It continuously updates its knowledge base with user feedback and discovered best practices for futu…
Orchestrates full software release lifecycle management
This skill orchestrates the entire software release lifecycle, managing everything from initial planning and dependency graph generation to execution scheduling and cross-feature integration testing. It supports complex workflows, enabling …
Triage, review, and merge open pull requests
This skill automates the end-to-end management of open pull requests. It assesses merge readiness by running security gates, checking CI status, managing required Opus reviews, and merging eligible changes while flagging release candidates …