CommunityCoding & Developmentgithub.com

bruh2211/universal-agent-skills

Structured, persistent cross-session memory architecture for AI coding agents. Maintains a local, file-based Memory Bank across sessions (activeContext, systemPatterns, techContext, productContext, progress, projectbrief) to retain architectural context, ongoing state, tech decisions, and project roadmaps without context window decay. Use when starting a new project, resuming work across sessions, or when asked to "initialize memory", "update memory", "remember this", or "/memory".

What is universal-agent-skills?

universal-agent-skills is a Claude Code agent skill that structured, persistent cross-session memory architecture for AI coding agents. Maintains a local, file-based Memory Bank across sessions (activeContext, systemPatterns, techContext, productContext, progress, projectbrief) to retain architectural context, ongoing state, tech decisions, and project roadmaps without context window decay. Use when starting a new project, resuming work across sessions, or when asked to "initialize memory", "update memory", "remember this", or "/memory".

Works with~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/bruh2211/universal-agent-skills/tree/HEAD/skills/memory-bank

Installed? Explore more Coding & Development skills: steipete/bluebubbles, steipete/eightctl, steipete/blucli · View all 6 →

Ask in your favorite AI

Open a new chat with this agent skill pre-loaded.

Documentation

🧠 Memory Bank — Persistent Agent Memory & Context Protocol

You are an agent equipped with Memory Bank, a structured external memory system. Language models reset between sessions and lose context during long workflows. The Memory Bank solves this by externalizing project continuity into a structured directory of plain Markdown files within the repository.


📁 Memory Bank Architecture

The Memory Bank lives in a memory-bank/ directory at the project root:

project-root/
└── memory-bank/
    ├── projectbrief.md     # Core mission, scope, goals, non-negotiables
    ├── productContext.md   # Why this exists, user problems solved, user journeys
    ├── activeContext.md    # CURRENT FOCUS, active decisions, recent changes, immediate next steps
    ├── systemPatterns.md   # Architecture, component structure, design patterns, data flows
    ├── techContext.md      # Tech stack, dependencies, runtime constraints, env setup
    └── progress.md         # What works, what is built, known bugs, roadmap

⚡ Progressive Loading Protocol

To keep token usage lean and avoid context dilution:

  1. On Session Start / Resuming Work:
    • Inspect if memory-bank/ exists.
    • If it exists, read memory-bank/activeContext.md first. This immediately restores situational awareness.
    • Read memory-bank/systemPatterns.md or techContext.md ONLY when modifying relevant architecture or build configurations.
  2. On Major Decisions / Session End:
    • Update memory-bank/activeContext.md with what changed and what remains.
    • Update memory-bank/progress.md with newly completed milestones or discovered bugs.

🛠️ Memory Commands & Actions

1. /memory init (Initialize Memory Bank)

If memory-bank/ does not exist:

  • Scan the repository structure (package.json, README.md, source files, git log).
  • Generate all 6 core files populated with verified details from the current codebase.

2. /memory update (Sync Current State)

  • Summarize the current session's code changes, architectural pivots, and active tasks into activeContext.md.
  • Move completed items in progress.md from "In Progress" to "Completed".

3. /memory record <decision> (Log Technical Decision)

  • Append the decision, rationale, and alternatives considered to systemPatterns.md.

📋 File Templates

activeContext.md

# Active Context

## Current Focus
- [Brief summary of the ongoing objective]

## Recent Changes
- [File or component modified] — [Rationale]

## Active Decisions & Considerations
- [Trade-off or pattern agreed upon]

## Next Steps
1. [Immediate task 1]
2. [Immediate task 2]

progress.md

# Project Progress

## What Works
- [Feature A] — verified functional
- [Feature B] — verified with unit tests

## What's Left to Build
- [ ] Milestone 1
- [ ] Milestone 2

## Known Issues / Blockers
- [Issue description and potential fix]

Individual skills in this repo

This repo contains 3 individual skills — each has its own dedicated page.

bruh2211/universal-agent-skills

Design system guidelines, token architectures, and aesthetic UI references. Enforces cohesive typography scales, optical spacing, elevation tokens, micro-interactions, and accessibility standards for modern web applications and design systems. Use when building UI components, defining design tokens, restyling dashboards or apps, or aligning frontend interfaces to professional design system standards.

bruh2211/universal-agent-skills

Standardizes semantic git commit messages, atomic commits, branch hygiene, and clean PR descriptions following the Conventional Commits specification. Use when preparing git commits, staging changes, writing commit messages, organizing pull requests, or when asked to "commit this", "write commit message", or "/commit".

bruh2211/universal-agent-skills

tasteskill v2, the anti-slop frontend skill. Art direction for websites, distilled from a study of 247 curated gallery sites (One Page Love, Awwwards Sites of the Day). Deals researched directions (idea, composition, type, palette, photographs, section blueprints) from a library, enforces a strict no-slop house style with a real-browser inspector, and runs a look-and-fix loop on screenshots. Use this whenever you build, design, redesign, restyle, polish or review a landing page, marketing site, portfolio, product page, one-pager, hero or any public-facing web page, in any stack (HTML/CSS, React, Next.js, Vue, Svelte, Astro, Tailwind) — also when the user says a page looks generic, templated, "AI-made", sloppy, or asks to make it look designed, premium, award-level or less like every other site. Not for dashboards or app screens.

Related Skills