Janitor for Openclaw

An automated maintenance agent designed to keep your AI workspace organized, healthy, and synchronized.

honeybee1130
v1.0.0
Feb 24, 2026
0
0
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install workspace-janitor

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install workspace-janitor 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 Janitor?

Janitor is a specialized background agent within the Openclaw Skills ecosystem dedicated to workspace hygiene and repository health. Its primary purpose is to act as a digital custodian, handling the repetitive but essential tasks of memory consolidation, document synchronization, and stale data removal. By maintaining the logical structure of a project, Janitor ensures that other agents have access to clean, relevant, and up-to-date information without the clutter of orphaned artifacts or outdated logs.

This skill is particularly effective for long-running projects where document drift occurs—situations where configuration files like TOOLS.md or status logs like HEARTBEAT.md no longer reflect the technical reality of the workspace. Janitor bridges this gap by performing routine checks and archiving old data, allowing the human developer to focus on high-level strategy while the Openclaw Skills framework handles the background upkeep.

Janitor Use Cases

  • Consolidating fragmented daily memory notes into a centralized MEMORY.md for better agent context.
  • Identifying and archiving stale research leads or data files older than 30 days to maintain focus.
  • Organizing messy artifact directories by moving orphaned or temporary files to dedicated archives.
  • Synchronizing system documentation and agent configs with the current workspace state during scheduled deep cleans.

How Janitor Works

  1. The agent triggers based on a scheduled interval (daily or weekly) or a specific maintenance command.
  2. It performs a comprehensive scan of the workspace against predefined checklists for memory health, artifact age, and document drift.
  3. Execution of cleanup protocols occurs, involving file moves to the archive directory and the distillation of learnings.
  4. A detailed maintenance report is generated, documenting every action taken, items flagged, and overall workspace health metrics.
  5. The report is saved to a standardized artifact path, and any items requiring human intervention are flagged for review.

Janitor Setup

To deploy the Janitor skill, ensure your environment is configured to allow file system management within the workspace. Create the necessary directory structure for logging:

mkdir -p /home/ubuntu/.openclaw/workspace/artifacts/janitor/

Integrate the Janitor configuration into your agent manifest. Ensure the agent has appropriate permissions to read from and write to the memory/ and research/ directories. For optimal results, schedule the weekly deep clean to run during low-activity periods to avoid interference with active Openclaw Skills tasks.

Janitor Data Schema & Taxonomy

Janitor follows a strict organizational taxonomy to ensure data remains recoverable even after cleanup:

Directory Data Type Retention Policy
memory/ Daily Notes Consolidated after 3 days
memory/archive/ Historical Context Archived after 7 days
artifacts/janitor/ Maintenance Logs Permanent history of actions
research/ Raw Data Flagged as stale after 30 days

All maintenance reports use the naming convention maintenance-[YYYY-MM-DD].md and include metadata on workspace health, such as active versus archived file counts.

Janitor Advanced Features

  • Distillation logic that extracts key learnings from daily logs into a permanent MEMORY.md file.
  • Automated flagging system that notifies Scout agents when research leads require re-validation.
  • Multi-directory orphan detection to find and clean up files generated by other automated Openclaw Skills.
  • Safety-first archival system that uses move-to-trash logic instead of permanent deletion for all content files.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*