← Back to Skills Marketplace
Self-Improving Engineering
by
José I. O.
· GitHub ↗
· v1.2.1
· MIT-0
100
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install self-improving-engineering
Description
Captures architecture decisions, code quality issues, build/deploy failures, dependency problems, performance regressions, tech debt accumulation, and test g...
Usage Guidance
This skill appears coherent and low-risk for its purpose, but review before enabling hooks: 1) Only enable the hooks you want (activator/UserPromptSubmit recommended). 2) Inspect scripts (activator.sh, error-detector.sh, extract-skill.sh) to confirm behavior and file paths. 3) If you enable PostToolUse/error-detector, be cautious because CLAUDE_TOOL_OUTPUT may contain sensitive command output—ensure the detector is configured to only emit short, redacted reminders. 4) Prefer installing from a trusted repository or vendor; if the source is unknown, review the full code locally before copying to ~/.openclaw or enabling hooks.
Capability Analysis
Type: OpenClaw Skill
Name: self-improving-engineering
Version: 1.2.1
The skill bundle is a legitimate engineering tool designed to help AI agents log and promote architectural decisions, build failures, and technical debt. It consists of markdown templates, shell scripts for scaffolding (extract-skill.sh), and OpenClaw hooks (handler.js/ts) that inject reminders into the agent's context. The scripts include basic security checks against path traversal, and the instructions explicitly advise the agent against logging sensitive information like secrets or private keys.
Capability Tags
Capability Assessment
Purpose & Capability
Name/description, README instructions, scripts, and hook handlers all implement a logging/prompting workflow to capture engineering learnings (.learnings, promotion to workspace docs). No unrelated credentials, binaries, or network endpoints are requested.
Instruction Scope
Runtime instructions create/ensure .learnings files and recommend installing optional hooks. Hook handlers inject a virtual reminder into agent bootstrap and shell scripts read CLAUDE_TOOL_OUTPUT for simple error-pattern detection. This is consistent with the skill's purpose, but the skill does write files to the workspace/home and (if PostToolUse is enabled) will read tool output—review to ensure no secrets or full stack traces are logged.
Install Mechanism
No automated install spec; install instructions are copy/clone or use clawdhub. All included scripts are local and there are no downloads from untrusted URLs or extract-from-URL steps.
Credentials
The skill declares no required environment variables or credentials. The only environment usage is reading CLAUDE_TOOL_OUTPUT in an optional error-detector script, which is appropriate for a PostToolUse hook but should be enabled only when you trust the runtime and its outputs.
Persistence & Privilege
always:false and hooks are user-enabled; handlers inject virtual files into session context (intended reminder). The skill does not request permanent platform-wide privileges or modify other skills' configs.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improving-engineering - After installation, invoke the skill by name or use
/self-improving-engineering - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.2.1
**Version 1.1.0**
- Added stackability contract for multi-skill installations.
- Added namespaced logging guidance (`.learnings/engineering/`) for coexistence with other skills.
- Added required `Skill: engineering` metadata field and cross-skill precedence/ownership rules.
- Clarified hook arbitration model (single dispatcher, dedupe, rate limiting).
v1.1.0
Version 1.1.0
- Switched to a manual-first workflow for logging engineering learnings and issues.
- Updated documentation to clarify that reminders/hooks are opt-in and not enabled by default.
- Adjusted hook instructions to emphasize lightweight, manual activation for reminders.
- No code or implementation changes; documentation improvements only.
v1.0.0
Initial release of self-improving-engineering skill for continuous engineering improvement.
- Provides structured logging of build failures, architecture decisions, code quality issues, performance regressions, dependency problems, tech debt, and test gaps.
- Introduces standardized markdown log files in a `.learnings/` directory: `LEARNINGS.md`, `ENGINEERING_ISSUES.md`, and `FEATURE_REQUESTS.md`.
- Offers detailed guidance for initializing log files and using the logging workflow across various engineering situations.
- Supports promotion of important learnings to architecture decision records, coding standards, or CI/CD runbooks.
- Includes integration recommendations for OpenClaw workspaces and generic setup instructions for other code agents.
Metadata
Frequently Asked Questions
What is Self-Improving Engineering?
Captures architecture decisions, code quality issues, build/deploy failures, dependency problems, performance regressions, tech debt accumulation, and test g... It is an AI Agent Skill for Claude Code / OpenClaw, with 100 downloads so far.
How do I install Self-Improving Engineering?
Run "/install self-improving-engineering" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self-Improving Engineering free?
Yes, Self-Improving Engineering is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Self-Improving Engineering support?
Self-Improving Engineering is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self-Improving Engineering?
It is built and maintained by José I. O. (@jose-compu); the current version is v1.2.1.
More Skills