proactive-agent-lite — Claude Code skill: setup and features
proactive-agent-lite turns passive agents into initiative-taking partners with memory preservation before compaction and automated self-healing.
MIT 105 installs ★ 121 updated 2026-11-03
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a claude-codeWhat proactive-agent-lite does
Instead of waiting for prompts, your agent anticipates workflow needs. The skill implements memory flushing right before window compaction occurs, which keeps critical project context intact across longer sessions.
It also applies reverse prompting to surface ideas and fixes internal errors automatically. You will need OpenClaw v1.0 or higher with a standard configuration to make it run.
What it helps with
- preserve mission context through pre-compaction flushes when the window fills
- surface unexpected ideas using reverse prompting
- diagnose and resolve internal agent errors automatically
- maintain strict alignment with user goals across tasks
Example prompts for the agent
- Suggest optimizations for this module that I did not explicitly request
- Diagnose recent failures and resolve the detected inconsistencies
- Flush key decisions into memory before the context window compacts
How to install proactive-agent-lite
1Skills CLI (Vercel)
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a claude-code2Manually: put the files in place
~/.claude/skills/proactive-agent-lite/SKILL.mdCopy the whole skill folder; for one project use .claude/skills/proactive-agent-lite/
1Skills CLI (Vercel)
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a codex2Manually: put the files in place
.agents/skills/proactive-agent-lite/SKILL.mdFor all projects use ~/.agents/skills/
1Skills CLI (Vercel)
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a cursor2Manually: put the files in place
.cursor/skills/proactive-agent-lite/SKILL.mdCursor also reads .agents/skills/ and .claude/skills/
1GitHub CLI
gh skill install aAAaqwq/AGI-Super-Team skills/proactive-agent-lite2Manually: put the files in place
.github/skills/proactive-agent-lite/SKILL.md1Skills CLI (Vercel)
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a gemini-cli2Platform CLI
gemini skills install https://github.com/aAAaqwq/AGI-Super-Team --path skills/proactive-agent-liteManually: put the files in place
.devin/skills/proactive-agent-lite/SKILL.mdThe legacy .windsurf/skills/ path still works
Skills CLI (Vercel)
npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a opencodeClaude.ai
Zip the folder and upload it in Settings → Capabilities → Skills
Step by step, with commands for each platform: How to install skills in Claude Code: five ways
Questions about proactive-agent-lite
How to install proactive-agent-lite in Claude Code?
Run npx skills add aAAaqwq/AGI-Super-Team --skill proactive-agent-lite -a claude-code in your terminal.
What does proactive-agent-lite do?
It enables unprompted suggestions, keeps context alive through pre-compaction memory flushes, and repairs internal inconsistencies automatically.
What are the dependencies for proactive-agent-lite?
It depends on OpenClaw v1.0 or higher alongside a standard agent configuration.
Source text
SKILL.md · MIT
---
name: proactive-agent-lite
description: Transform AI agents from task-followers into proactive partners with
memory architecture, reverse prompting, and self-healing patterns. Lightweight version
focused on core proactive capabilities.
metadata:
openclaw:
requires: {}
install: []
author: Daniel Li
---
# Proactive Agent Lite
- Author: Daniel Li
- Copyright © Daniel Li. All rights reserved.
Transform your AI agents from passive task-followers into proactive partners that anticipate needs and continuously improve.
## Core Features
- **Memory Architecture**: Pre-compaction flush ensures context survives when window fills
- **Reverse Prompting**: Surfaces ideas you didn't know to ask for
- **Security Hardening**: Built-in security considerations and safe defaults
- **Self-Healing Patterns**: Diagnoses and fixes its own issues automatically
- **Alignment Systems**: Stays on mission and remembers who it serves
## Key Benefits
✅ **Anticipates Needs**: Proactively suggests solutions before you ask
✅ **Continuous Learning**: Improves from every interaction without explicit training
✅ **Mission-Focused**: Never loses sight of the core objective
✅ **Self-Maintaining**: Automatically recovers from errors and inconsistencies
✅ **Value Creation**: Generates insights and opportunities without being prompted
## Usage
This skill enhances any OpenClaw agent by providing proactive behavior patterns. Simply install and the agent will automatically begin exhibiting proactive characteristics.
## Integration
Works seamlessly with all OpenClaw agents and can be combined with other skills for enhanced functionality.
## Requirements
- OpenClaw v1.0 or higher
- Standard agent configuration
## Customization
The proactive behavior can be tuned through configuration parameters to match your preferred level of initiative and communication style.
© Daniel Li · MIT · aAAaqwq/AGI-Super-Team