A universal skill providing agents with a domain-filtered, 14-day rolling context and priority management system.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install hippocampus-sync
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install hippocampus-sync using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is Hippocampus Sync?
Hippocampus Sync is a specialized utility within the Openclaw Skills framework designed to maintain a high-fidelity memory state for AI agents. It functions by incrementally updating a HIPPOCAMPUS.md file, which serves as the agent's primary decision-making context. By synthesizing data from activity feeds, local memory files, and peer agent states, it ensures that agents stay focused on their specific domain while maintaining awareness of broader system activities.
The skill enforces strict domain boundaries through a registry-based tracking system. This prevents information bloat by compressing peer data into single-line summaries while providing deep, high-fidelity insights into the agent's own workspace. This structured approach to memory allows for more efficient processing and more accurate decision-making across complex multi-agent environments.
Hippocampus Sync Use Cases
- Automating the daily synchronization of agent priorities and Top of Mind items.
- Maintaining a rolling 14-day history of recent sessions and outcomes to inform current tasks.
- Tracking persistent commitments and pending decisions that must survive beyond the standard memory window.
- Synchronizing cross-workspace context while preventing domain bleed between specialized agents.
How Hippocampus Sync Works
- Identify the active agent via the IDENTITY.md file and look up its configuration in the central registry.
- Query the authoritative events.db activity feed for the last 24 hours of data.
- Retrieve the last 14 days of agent-specific memory files to build a baseline context.
- Apply domain-specific Track and Exclude filters to ensure the information processed is relevant to the agent's role.
- Compress information from peer agents into single-line directional context.
- Refresh the Top of Mind section with 3-5 high-priority items weighted by the agent's specific voice and urgency criteria.
- Update persistent sections including Open Threads, Commitments, and Recent Sessions.
- Validate all dates and entries against source files to ensure technical accuracy.
- Execute a learnings decay check to archive stale data and flag files requiring manual compaction.
- Write the finalized HIPPOCAMPUS.md and append material changes to the agent's memory logs.
Hippocampus Sync Setup
To initialize the skill within your Openclaw Skills environment, ensure your agent registry and domain rules are defined in the references directory. Run the following command to verify the data decay script:
python3 ~/.openclaw/skills/hippocampus-sync/scripts/decay.py "$(pwd)"
Ensure that the agent has read access to ~/.openclaw/events.db and the appropriate references/agent-registry.md file.
Hippocampus Sync Data Schema & Taxonomy
The skill manages data across several key files and formats to maintain high context density:
| File/Source | Role | Retention Policy |
|---|---|---|
| HIPPOCAMPUS.md | Primary context file | 14-day rolling for sessions; infinite for threads/commitments |
| events.db | Authoritative activity feed | Ground truth for the last 24 hours |
| memory/YYYY-MM-DD.md | Daily activity logs | Used for incremental sync over 14 days |
| .learnings/archive/ | Long-term data storage | Archived after 90 days of staleness |
The HIPPOCAMPUS.md file uses a structured Markdown hierarchy including Top of Mind, Open Threads, Commitments, and a Recent Sessions table.
Hippocampus Sync Advanced Features
- Context Compression: Automatically reduces peer agent data to single lines to maintain focus.
- Dynamic Weighting: Prioritizes information differently based on agent types, such as severity for diagnostic agents or revenue impact for advisory agents.
- Automated Decay: A Python-based script that automatically marks pending entries as stale after 30 days and archives them after 90 days.
- Domain Enforcement: A secondary verification step that removes any items falling under the Exclude list, regardless of the source file content.
- Multi-Agent Support: Universal application across all Openclaw Skills agents using a centralized registry.
SKILL.md
Loading
Related Openclaw Skills
Seedance 2.0 Prompt Engineering to Video
A specialized tool for designing production-grade English prompts and generating high-control AI videos using the Seedance 2.0 model on WeryAI.
Agent Briefing — Not For Humans
A zero-configuration skill for monitoring daily AI agent briefings and extracting structured product compatibility scores from the Not For Humans channel.
LLM Performance Estimator
A technical tool for calculating Time To First Token (TTFT), token generation speed, and VRAM footprints for Large Language Models on specific hardware.
OpenClaw Model Switcher
A specialized WebUI and CLI utility for rapid switching and management of AI model providers within the OpenClaw ecosystem.
Buddhism & OpenClaw Ecosystem
A bilingual knowledge skill offering deep insights into Buddhist philosophy and the technical project matrix of the OpenClaw ecosystem.
Roon Music Controller
A sophisticated AI agent skill for full-featured control of Roon music systems including library search, zone management, and playback automation.
Featured*
