cy-workflow-memory

Read and update workflow memory files across Compozy PRD task runs.

2.5k|134|Updated Mar 28, 2026
One-click install
npx skills add https://github.com/compozy/compozy --skill cy-workflow-memory-compozy
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: cy-workflow-memory
Source: https://github.com/compozy/compozy/tree/main/.agents/skills/cy-workflow-memory
Command: npx skills add https://github.com/compozy/compozy --skill cy-workflow-memory-compozy

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Maintains workflow-scoped task memory for Compozy runs using .compozy/tasks/<name>/memory/ files. Use when a task prompt provides workflow memory paths and requires the agent to read, update, compact, and promote durable context across PRD task executions. Do not use for PR review remediation, global user preferences, or programmatic event-log summarization.

Core Features & Use Cases

  • Reads and writes the shared workflow memory to carry cross-task context across PRDs and runs.
  • Keeps task-local memory separate from shared memory, promoting only durable learnings into the shared memory.
  • Uses references/memory-guidelines.md to guide compaction decisions and ensure clean handoffs.

Quick Start

Load the shared workflow memory and the current task memory, update memory as the task objective evolves, and compact when requested to promote durable cross-task context.

Frequently Asked Questions about cy-workflow-memory

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

FAQPage Schema
How do I maintain durable workflow memory across multiple AI agent tasks?

Durable workflow memory is maintained by reading and updating memory files within a shared directory structure. This allows agents to carry cross-task context across PRD task runs and compacts context to ensure clean handoffs.

What is cross-task memory compaction for long-running AI pipelines?

Cross-task memory compaction is the process of summarizing and promoting durable learnings into shared memory. It uses specific guidelines to guide compaction decisions, ensuring clean context handoffs between subsequent task executions.

How do I separate task-local memory from shared workflow memory?

You separate task-local memory from shared workflow memory by keeping local memory files for quick reuse and promoting only durable cross-task learnings into the shared memory state. This prevents context bloat across multiple runs.

When should I not use cross-task workflow memory management?

Cross-task workflow memory management should not be used for PR review remediation, global user preferences, or programmatic event-log summarization. It is specifically designed for long-running, multi-task Compozy pipelines.

Can I use workflow memory to carry context across different PRD task runs?

Yes, workflow memory carries cross-task context across PRD task runs and across separate runs. It reads and writes to a shared workflow memory path to maintain state for long-running AI-assisted workflows.

How do I compact task memory to ensure clean handoffs between agents?

You compact task memory by applying compaction guidelines from a references file to promote durable cross-task context. This keeps task memory local for quick reuse while ensuring clean handoffs for the next agent.