pson-agent

Collect user signals with PSON5 and project redaction-aware agent context.

4|Updated Apr 22, 2026
One-click install
npx skills add https://github.com/fredabila/pson5 --skill pson-agent
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: pson-agent
Source: https://github.com/fredabila/pson5/tree/main/skills/pson-agent
Command: npx skills add https://github.com/fredabila/pson5 --skill pson-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Collects user signals, stores them in a structured profile, and projects an agent-safe context so AI agents can personalize behavior with privacy and consent safeguards.

Core Features & Use Cases

  • End-to-end personalization pipeline: acquisition → modeling → state → graph → simulation → projection.
  • Agent-safe projection with consent, redaction, and trust boundaries to prevent leakage of sensitive data.
  • Multi-interface integration: works with the SDK, HTTP API, CLI, or MCP for end-to-end personalization in diverse environments.

Quick Start

Install the SDK, create a user profile, fetch agent context for a target intent, and then use the projection to tailor the response.

Frequently Asked Questions about pson-agent

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

FAQPage Schema
How do I personalize AI agent behavior using structured user profiles?

AI agent personalization collects user signals into a structured profile and projects an agent-safe context. This projected context is then passed to the AI agent to tailor its behavior across diverse interfaces while maintaining privacy safeguards.

What is the best way to prevent sensitive data leakage in AI agent personalization?

Agent-safe projection prevents sensitive data leakage by applying consent rules, redaction, and trust boundaries during context projection. This ensures only safe, filtered personalization data is exposed to downstream AI components.

How do I integrate agent personalization across SDK, HTTP API, CLI, and MCP environments?

Multi-interface integration allows you to deploy agent personalization across SDK, HTTP API, CLI, or MCP environments. This enables end-to-end personalization workflows in diverse technical stacks without requiring custom adapters for each interface.

Can I use PSON5 structured data to manage AI agent memory and consent rules?

PSON5 structured data manages AI agent memory and consent rules by storing user signals in a structured profile. The system applies consent and privacy rules during projection to ensure compliant behavior modeling and safe context exposure.

Does agent personalization support registry workflows and provider adapters?

Agent personalization supports registry workflows, generative workflows, and provider adapters. It utilizes seven tools with atomic writes to ensure reliable state management and seamless integration across different AI provider environments.

What is the end-to-end pipeline for AI agent personalization and simulation?

The end-to-end personalization pipeline covers acquisition, modeling, state, graph, simulation, and projection. This structured process ensures user signals are safely transformed into agent-safe context for tailored AI behavior.