Skill UI

This skill provides structured patterns for combining vector similarity search with traditional keyword-based search. It is essential for building advanced Retrieval-Augmented Generation (RAG) systems, sophisticated search engines, or any retrieval task requiring high recall. Use this when pure semantic understanding is insufficient, and precise matching of specific terms (like names or codes) is necessary.

RecallMax dramatically enhances AI agent memory capabilities by cleanly injecting massive external contexts (500K-1M tokens) without hallucination drift. It automatically summarizes lengthy conversations, preserving tone, intent, and key facts. Furthermore, it compresses multi-turn interaction histories into high-density token sequences, ensuring agents maintain perfect recall across complex and long-running sessions.

Describes how to use the MemOS local memory toolkit to recall past conversations, search shared histories, and choose the right memory tool when automatic recall is insufficient.

EverOS OpenClaw Plugin integrates persistent, long-term memory capabilities into OpenClaw using the ContextEngine API. It goes beyond simple memory slots by automatically recalling relevant past context before a reply and saving new conversation content after a turn. This allows users to chat naturally while ensuring context, preferences, and key details persist across multiple sessions, eliminating the need for manual memory calls.

A comprehensive skill designed to guide users through the installation and configuration of the OpenViking long-term memory plugin for OpenClaw. It enables the agent to automatically capture facts from conversations and recall relevant context across sessions (auto-capture + auto-recall). The process is fully natural-language driven, requiring no command-line knowledge from the user.

A self-hosted, semantic memory service designed for AI agents. It allows agents to save worklogs, decisions, and notes, and recall this knowledge across sessions based on meaning rather than exact keywords. Using pgvector and structured workspaces, it maintains persistent, context-aware knowledge bases, making it ideal for collaborative AI projects.

Query and manage AgentDB knowledge base using controller bridge tools. Supports semantic routing, hierarchical recall, causal graphs, and context synthesis for agent sessions. Enables storing, retrieving, and searching structured knowledge across tasks.

This powerful skill leverages learned patterns and the current operational state to predict the most efficient next action. It guides complex workflows by offering immediate, high-confidence predictions. If confidence is low, it intelligently falls back to task priority lists. After every completed task, it utilizes a robust learning loop to update success patterns, ensuring continuous process improvement and cross-session memory recall.

This operator skill provides comprehensive control over the OpenViking context engine, enabling agents to manage long-term memory, session archives, and external resources within OpenClaw. It facilitates semantic recall, context querying, resource import, and recall trace debugging by communicating with a remote OpenViking server. Use this plugin when your agent requires advanced, persistent context management.

This skill audits OpenMed NER or de-identification models to detect performance disparities across sensitive demographic subgroups (e.g., sex, age, race). It calculates metrics like per-group recall and leakage disparity, helping users identify instances of 'under-protection' before deployment, ensuring equitable AI governance and compliance.

This tool benchmarks the recall of PII detection models, crucial for ensuring privacy compliance before deployment. It uses synthetic fixtures and enforces a recall floor by measuring label-aware exact-span and grapheme recall against predefined gold standards. It is essential for auditing models and preventing critical privacy leaks.

This skill provides an in-depth, entity-level scorecard for clinical or biomedical Named Entity Recognition (NER) models. It calculates precision, recall, and F1 scores (both strict and partial match modes) using a user-supplied gold corpus. Key features include per-label error breakdown, a confusion matrix, and detailed False Negative/False Positive analysis, essential for debugging model performance in highly sensitive medical domains.