Skill UI

The Ejentum Reasoning Harness is a comprehensive library of cognitive operations designed to guide complex AI agents through critical thinking tasks. It exposes four specialized harnesses (reasoning, code, anti-deception, memory) that mitigate common failure modes like hallucination and reasoning decay. Instead of a fixed answer, the agent ingests a detailed, executable scaffold—including failure patterns, procedures, and falsification tests—to ensure robust, self-verified output.

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.

This skill captures and persists critical user decisions, project constraints, and architectural choices across multiple sessions and agents. It saves this context into a structured memory database (`mdbase`), ensuring that future development steps or agents can retrieve the original intent without re-deriving it from scratch. Ideal for documenting project rules, tech stacks, and design rationale.

A meta-skill designed to monitor long-running processes, such as model fine-tunes or CI builds, executed within a detached tmux session. It captures output, automatically uses a sub-agent to diagnose failures, proposes a fix, and records the complete debugging process to memory. Ideal for intermittent failure debugging and continuous integration pipelines.

SkillOpt-Sleep provides a mechanism for autonomous agent self-improvement. By simulating a 'sleep' or 'dream' cycle, it harvests past interaction sessions, mines recurring tasks, and replays them through the designated backend. It consolidates learned knowledge into persistent memory and skills, ensuring that only strictly validated improvements are adopted via a held-out gate, simulating long-term competence gain from short-term experience. Ideal for scheduled, background optimization.

This skill provides a comprehensive guide for building scalable Retrieval-Augmented Generation (RAG) pipelines and persistent agent memory using Pinecone as the vector database. It covers index setup, embedding large corpuses, implementing dense and hybrid semantic search, and advanced metadata filtering for real-world production systems. Use this for building searchable knowledge bases and context-aware AI agents.

This skill enables the creation of a robust 'harness' system for coding agents within a repository. It addresses the need to prevent repeated AI-generated coding mistakes by implementing durable instructions, guardrails, regression tests, and failure memory. It treats the target repository as the source of truth, ensuring consistency and maintaining high reliability across all automated contributions.

This skill provides a hybrid, persistent, and searchable knowledge memory system specifically designed for enhancing AI agents' capabilities. It allows developers to record, retrieve, and search key knowledge, architectural decisions, and operational memories automatically. Use it to build agents with long-term context and reliable knowledge management, ensuring findings are synced back into project documentation.

Context Kit is an external framework designed to organize and manage personal, durable context data (like identity details, memories, and domain knowledge) into secure Markdown artifacts. It helps users set up robust knowledge bases for coding agents (e.g., Claude Code), guiding them on best practices for data privacy, local storage, and safe integration, while preventing sensitive data leaks.

Diagnoses full-stack failure modes in agent and LLM applications, including wrapper regressions, memory contamination, tool discipline failures, hidden repair loops, and rendering corruption. Produces severity-ranked findings and code-first fixes for production agent systems.

Design a persistent multi-agent operating system on Claude Code with kernel routing, specialist agents, slash commands, file-based memory, and scheduled automation for long-lived projects.

This harness transforms standard code environments into fully persistent, autonomous agent systems. It leverages native features like scheduled tasks (Crons), remote dispatching, and structured memory storage (MCP) to enable continuous, self-directing automation. Ideal for building AI assistants that require long-term context, recurring workflows, and complex, multi-step operations, eliminating the need for external agent frameworks like AutoGPT.