para-memory-files

Store, update, and retrieve durable knowledge using the PARA method.

Updated Mar 15, 2026
One-click install
npx skills add https://github.com/mighty-hq/mighty-paperclip --skill para-memory-files-mighty-hq
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: para-memory-files
Source: https://github.com/mighty-hq/mighty-paperclip/tree/main/skills/para-memory-files
Command: npx skills add https://github.com/mighty-hq/mighty-paperclip --skill para-memory-files-mighty-hq

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This skill solves the problem of losing useful context between sessions by providing a durable, structured file-based memory that can be written to and recalled reliably over time.

Core Features & Use Cases

  • PARA knowledge graph for durable facts: Store atomic, entity-based YAML facts under Projects, Areas, Resources, and Archives, and keep entity summaries current.
  • Daily notes for a raw timeline: Capture the “when” layer continuously in dated Markdown files for conversational continuity and event history.
  • Tacit knowledge for user patterns: Record preferences and operating lessons in a dedicated MEMORY.md so the system learns how the user behaves.
  • Memory recall via qmd: Retrieve relevant context using semantic search, keyword search, or vector search, indexed over the agent home folder.
  • Planning support: Keep shared plans in a plans/ area so multiple agents can access up-to-date objectives.

Quick Start

Use para-memory-files to remember something new by writing it to the appropriate file in $AGENT_HOME, like updating the daily note or the relevant PARA entity folder.

Frequently Asked Questions about para-memory-files

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

FAQPage Schema
How do I persist context and recall knowledge across agent sessions?▼

You can persist and recall knowledge across sessions by writing atomic facts to a structured PARA file system under $AGENT_HOME, then retrieving them through qmd semantic and keyword search queries.

What does the PARA method do for knowledge management and daily notes?▼

The PARA method organizes durable knowledge into Projects, Areas, Resources, and Archives, while separate daily notes capture a raw timeline of events to maintain conversational continuity over time.

How do I store atomic facts and update entity summaries in YAML?▼

Store atomic facts by writing entity-based YAML files into the appropriate PARA category folder, ensuring entity summaries are kept current to maintain an accurate knowledge graph for retrieval.

Can I use this approach to manage shared plans with multiple agents?▼

Yes, you can maintain shared plans by keeping objectives in a dedicated plans area within the agent home folder, allowing multiple agents to access and synchronize up-to-date planning management files.

How does memory decay and recency work with summary-driven synthesis?▼

Memory decay applies summary-driven recency by synthesizing weekly timelines, allowing the system to compress older daily notes into summaries while prioritizing the most recently updated knowledge.

Do I need a specific database to run semantic search over my knowledge files?▼

No database is required; semantic search and recall are executed through qmd indexing directly over the markdown and YAML files stored persistently in your local agent home directory.