Below you will find pages that utilize the taxonomy term “Hjs-Spec”
April 18, 2026
Agent-Blackbox Blame Finder
Version updated for https://github.com/hjs-spec/Agent-Blackbox to version v0.1.0.
This action is used across all versions by ? repositories. Action Type This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary Agent Blame-Finder is a GitHub Action designed to automate fault analysis in multi-agent systems by tracking and identifying the root cause of failures. It installs a “black box” in each agent to record decision-making processes, their dependencies, timestamps, and cryptographic signatures to ensure accountability. The action provides instant blame analysis, visualizes causality chains, and highlights problematic agents, saving time and effort in troubleshooting complex systems.
April 16, 2026
JEP Accountability for GitHub
Version updated for https://github.com/hjs-spec/jep-github-action to version v0.1.0.
This action is used across all versions by ? repositories. Action Type This is a Node action using Node version 20.
Go to the GitHub Marketplace to find the latest changes.
Action Summary The “JEP Accountability for GitHub” action automates the generation of cryptographically signed Judgment Event Protocol (JEP) accountability receipts for AI or automated operations on GitHub, such as pull request merges, creations, and issue closures. It provides transparency by posting receipts as comments, adding commit status checks, and optionally sending events to a centralized API for storage. This action is designed to enhance accountability and traceability for automated workflows involving bots or AI agents.
April 16, 2026
shutup - MCP Tool Filter
Version updated for https://github.com/hjs-spec/shutup-mcp to version v0.2.0.
This action is used across all versions by ? repositories. Action Type This is a Composite action.
Go to the GitHub Marketplace to find the latest changes.
Action Summary shutup is a GitHub Action designed to streamline the interaction between MCP (Model Context Protocol) agents and servers by filtering and presenting only the most relevant tools based on the user’s intent. It reduces token usage, improves tool selection accuracy, and accelerates response times by aggregating tool lists across multiple MCP servers, creating a dynamic, privacy-preserving index, and limiting visible tools to the top 3-5 matches. This action automates tool discovery, ensures efficient context management, and enhances the agent’s decision-making process.