agent-context-keeper

Apply delta changes to .agent/ context on code edit triggers.

4|Updated Mar 31, 2026
One-click install
npx skills add https://github.com/ryan-nguyen-01/agent-platform --skill agent-context-keeper
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: agent-context-keeper
Source: https://github.com/ryan-nguyen-01/agent-platform/tree/main/.claude/Agents/agent-context-keeper
Command: npx skills add https://github.com/ryan-nguyen-01/agent-platform --skill agent-context-keeper

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Agent-context-keeper maintains an accurate and up-to-date .agent/ context across all agents by performing delta-based synchronization, resolving conflicts, and managing dirty flags, instead of rebuilding the entire state.

Core Features & Use Cases

  • Delta-based synchronization minimizes work by applying only changed sections.
  • Conflict resolution and locking prevent race conditions when multiple agents update context.
  • Dirty-flag driven workflow with triggers, changelogs, and audit trails keeps agents aligned.
  • On-demand activation by orchestrator or manual trigger after major changes.

Quick Start

Trigger a context sync by modifying code in the repository to observe the delta-based update flow and updated .agent/ context.

Frequently Asked Questions about agent-context-keeper

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

FAQPage Schema
How do I keep agent context in sync after code edits?

Delta-based synchronization updates the agent context by applying only changed sections triggered by file saves, ensuring all agents operate on current state without rebuilding the entire context.

How does delta sync prevent race conditions in a multi-agent orchestrator?

Conflict resolution and locking mechanisms prevent race conditions by enforcing frontmatter metadata, token budgets, and dirty-flag lifecycles when multiple agents update context simultaneously.

What triggers a context refresh in a multi-agent orchestrator?

Context refresh is triggered by commit, pull, file-save, and manual orchestrator triggers, applying delta changes to maintain timely and accurate agent state updates.

Can I audit context updates when multiple agents modify shared state?

Dirty-flag driven workflows with changelogs and audit trails ensure all context updates are auditable, keeping agents aligned and enforcing frontmatter metadata during synchronization.

What is the best way to update .agent context without rebuilding the entire state?

Delta-based synchronization minimizes work by applying only changed sections to the .agent context, enforcing token budgets and frontmatter metadata for safe, auditable updates.

When should I avoid full context rebuilds in favor of delta synchronization?

Full context rebuilds are unnecessary when delta sync can apply changed sections from code edits, managing dirty flags and conflicts to ensure agents operate on current state safely.