intertrust

Track AI agent trust scores with severity-weighted decay in SQLite.

3|1|Updated Feb 15, 2026
One-click install
npx skills add https://github.com/mistakeknot/Demarch --skill intertrust
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: intertrust
Source: https://github.com/mistakeknot/Demarch/tree/main/.gemini/generated-skills/intertrust
Command: npx skills add https://github.com/mistakeknot/Demarch --skill intertrust

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) and references (resource) components.

What problem does it solve?

This Skill addresses the challenge of managing and evaluating the reliability of AI agents by providing a robust system for tracking their performance and reputation.

Core Features & Use Cases

  • Reputation Tracking: Monitors agent performance through feedback on findings and their severity.
  • Severity-Weighted Decay: Applies a decay mechanism to trust scores, giving more weight to recent and severe findings.
  • Suppression Candidates: Identifies agents that may require closer monitoring or intervention.
  • Use Case: In a multi-agent development environment, this Skill helps identify which agents are consistently producing high-quality, accepted code versus those that frequently introduce issues, allowing for targeted improvements or reassignments.

Quick Start

Check the current trust status for all agents in the 'my-project' project.

Frequently Asked Questions about intertrust

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I track AI agent reputation and trust scores in a development lifecycle?

Agent trust scoring tracks AI agent reputation by storing feedback in a SQLite database, applying a blended algorithm with severity-weighted decay to evaluate historical performance and reliability.

What is severity-weighted decay in agent trust scoring?

Severity-weighted decay is a scoring mechanism that applies time-based decay to trust scores, giving more weight to recent and severe findings to evaluate agent reliability.

How do I identify AI agents that need closer monitoring or intervention?

Suppression candidates are identified by analyzing historical performance data and trust feedback, flagging agents that frequently introduce issues for targeted improvements or reassignments.

Can I use a SQLite database for storing multi-agent trust feedback?

Yes, agent trust scoring utilizes a shared SQLite database for storing trust feedback, calculating project-specific and global scores using a blended algorithm with time-based decay.

Does agent trust scoring evaluate both project-specific and global performance?

Yes, the trust scoring algorithm blends project-specific and global scores, incorporating time-based decay to provide a comprehensive evaluation of agent performance across environments.

When do I need agent trust scoring for my development workflow?

Agent trust scoring is needed in multi-agent development environments to identify which agents consistently produce accepted code versus those that frequently introduce issues.