← Back to Skills Marketplace
Self-Improving Domotics
by
José I. O.
· GitHub ↗
· v1.0.0
· MIT-0
79
Downloads
0
Stars
0
Active Installs
1
Versions
Install in OpenClaw
/install self-improving-domotics
Description
Captures smart-home automation conflicts, sensor drift, device connectivity failures, integration regressions, safety rule gaps, and energy optimization oppo...
Usage Guidance
This skill appears coherent and documentation-first, but review before enabling hooks: (1) inspect and approve the hook files (hooks/openclaw/*) and scripts (scripts/*.sh) — they run locally and will create files under your workspace; (2) be cautious enabling the PostToolUse hook: error-detector.sh inspects tool output (CLAUDE_TOOL_OUTPUT) which may contain sensitive info — the script prints reminders locally but does not send data off-host; (3) follow the skill's own guidance and avoid logging secrets (PINs, alarm codes, network credentials) into .learnings files; (4) if you plan to use the suggested git clone, verify the remote repository and review files locally before execution. If you want extra assurance, run the scripts with --dry-run or review them line-by-line in a safe environment before enabling hooks.
Capability Analysis
Type: OpenClaw Skill
Name: self-improving-domotics
Version: 1.0.0
The skill bundle is a documentation and logging framework designed to track smart-home (domotics) automation issues, sensor drift, and safety gaps. It utilizes shell scripts (activator.sh, error-detector.sh) and OpenClaw hooks (handler.js) to provide reminders and scaffold new log entries in a .learnings/ directory. The code is transparent, lacks any network exfiltration or obfuscation, and includes proactive security measures such as path-traversal checks in extract-skill.sh and explicit instructions to avoid logging sensitive credentials or secrets.
Capability Assessment
Purpose & Capability
The name/description match the included artifacts: markdown templates, scaffolding script, reminder/scan hooks, and examples for logging domotics learnings and issues. The extract-skill scaffold, activator, and hook handlers serve the stated purpose of creating and injecting reminder content and scaffolds.
Instruction Scope
Runtime instructions and scripts are reminder/documentation-only and explicitly warn not to actuate devices or log secrets. Scripts create/append files under .learnings/ or ./skills which is expected. One point to note: scripts/error-detector.sh reads the CLAUDE_TOOL_OUTPUT environment variable (tool output) to scan for patterns — this is not declared in requires.env but is a plausible runtime context provided by the hosting agent. Enabling the PostToolUse hook means the script will inspect local tool outputs, which can contain sensitive data; the skill does not transmit that data off-host.
Install Mechanism
There is no packaged install spec — the README suggests either clawdhub install or a git clone. All code is included in the repo; there are no external downloads, URL shorteners, or archive extraction steps in the install workflow. Scripts run locally and create files in user/workspace directories as expected.
Credentials
The skill requests no credentials or secret environment variables. The only runtime environment dependency visible in code is CLAUDE_TOOL_OUTPUT (used by error-detector.sh) and standard OpenClaw hook event fields (sessionKey, context). These are reasonable for hooks, but CLAUDE_TOOL_OUTPUT may contain sensitive tool output — the skill reads it locally (for pattern detection) but does not declare it as a required env var.
Persistence & Privilege
always is false and the skill is user-invocable. Optional hooks are enabled only when the user copies/enables them; the handlers only inject virtual reminder files into bootstrap context and do not modify other skills or system-wide configs. The skill does write scaffolds into relative ./skills or .learnings directories when used, which aligns with its stated behavior.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improving-domotics - After installation, invoke the skill by name or use
/self-improving-domotics - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
- Initial release of the self-improving-domotics skill for smart-home automation quality tracking.
- Provides structured guidance for logging learnings, issues, and feature requests to markdown files for continuous domotics improvement.
- Detects and captures key automation challenges: conflicts, sensor drift, device connectivity, integration breaks, safety gaps, latency, and energy optimization.
- Emphasizes documentation-only usage; does not execute real-world actions.
- Includes setup instructions for OpenClaw and generic agent environments.
- Supplies detailed templates and best practices for safe, clear incident and improvement logging.
Metadata
Frequently Asked Questions
What is Self-Improving Domotics?
Captures smart-home automation conflicts, sensor drift, device connectivity failures, integration regressions, safety rule gaps, and energy optimization oppo... It is an AI Agent Skill for Claude Code / OpenClaw, with 79 downloads so far.
How do I install Self-Improving Domotics?
Run "/install self-improving-domotics" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self-Improving Domotics free?
Yes, Self-Improving Domotics is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Self-Improving Domotics support?
Self-Improving Domotics is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self-Improving Domotics?
It is built and maintained by José I. O. (@jose-compu); the current version is v1.0.0.
More Skills