chain-link

Save a work checkpoint to preserve context for multi-session projects. Use when pausing work on a feature, completing a milestone, or before ending a session.

⭐3stars🔀0forks📁View on GitHub🕐Updated Jan 5, 2026

When & Why to Use This Skill

Chain Link is a specialized context preservation tool designed to bridge the gap between multiple AI sessions. It enables users to save structured work checkpoints—including technical decisions, code changes, and pending tasks—ensuring that complex projects maintain continuity and precision even when work is paused or resumed across different sessions.

Use Cases

  • Multi-session Development: Save a comprehensive snapshot of your current progress when pausing work on a complex feature to ensure the AI retains full context and technical nuances in the next session.
  • Milestone Documentation: Automatically generate and store structured summaries of completed tasks, file modifications, and architectural decisions upon reaching a project milestone for better version tracking.
  • Session Handoff and Continuity: Capture unresolved blockers and prioritized 'next steps' before ending a session, allowing for an immediate and productive start without needing to re-explain the project state.
namechain-link
descriptionSave a work checkpoint to preserve context for multi-session projects. Use when pausing work on a feature, completing a milestone, or before ending a session.

Chain Link - Save Work Checkpoint

Save your current work context as a chain link for later retrieval.

Usage

/chain-link [chain-name] [slug]

Instructions

  1. Gather Context - Analyze the current session:

    • What was accomplished
    • Key technical decisions made
    • Code changes with file paths
    • Unresolved issues or blockers
    • Next steps
  2. Generate Chain Content - Structure as XML:

<chain-link>
  <summary>Brief overview of work done</summary>

  <completed>
    <item>Task or feature completed</item>
    <item>Another completed item</item>
  </completed>

  <code-changes>
    <file path="src/auth.rs">Added JWT validation middleware</file>
    <file path="src/routes.rs">New /login endpoint</file>
  </code-changes>

  <decisions>
    <decision reason="Performance">Used Redis for session storage</decision>
    <decision reason="Security">JWT tokens expire after 1 hour</decision>
  </decisions>

  <issues>
    <issue status="blocked">Rate limiting not implemented yet</issue>
    <issue status="investigation">Memory leak in connection pool</issue>
  </issues>

  <next-steps>
    <step priority="high">Implement rate limiting middleware</step>
    <step priority="medium">Add refresh token support</step>
  </next-steps>
</chain-link>
  1. Save via MCP - Call tinymem_chain_link:
    • session_id: from TINYMEM_SESSION env
    • chain_name: provided or infer from task context
    • slug: provided or generate from summary (kebab-case)
    • content: XML structured content above

Examples

User: /chain-link auth-feature jwt-middleware → Saves checkpoint with chain "auth-feature", slug "jwt-middleware"

User: /chain-link → Ask user for chain name, auto-generate slug from work summary