← Back to Skills Marketplace
Self-Improving Support
by
José I. O.
· GitHub ↗
· v1.1.0
· MIT-0
94
Downloads
0
Stars
0
Active Installs
2
Versions
Install in OpenClaw
/install self-improving-support
Description
Captures ticket resolution delays, misdiagnoses, escalation gaps, SLA breaches, knowledge gaps, and customer churn signals to enable continuous support impro...
Usage Guidance
This skill appears coherent for support-process improvement. Before enabling: (1) review the scripts (activator.sh, error-detector.sh, extract-skill.sh) to confirm you’re comfortable with local file writes (/.learnings/ or ./skills/) and that no PII will be logged; (2) only enable the hooks you need—UserPromptSubmit (activator) is lightweight; PostToolUse (error detector) reads tool output and should be enabled only if you trust automated inspection of command outputs; (3) if you install manually, prefer reviewing the GitHub repo contents first; (4) consider adding a matcher in your hook configuration so reminders run only for support-related prompts; and (5) remember the skill can’t enforce anonymization—agents must follow the “never log PII” guidance.
Capability Analysis
Type: OpenClaw Skill
Name: self-improving-support
Version: 1.1.0
The self-improving-support skill is designed to help support agents track ticket issues, knowledge gaps, and feature requests. The bundle contains markdown templates, shell scripts for initialization and skill scaffolding (extract-skill.sh), and OpenClaw hooks (handler.js/ts) that inject reminders into the agent's context. Analysis of the code shows no evidence of data exfiltration, malicious execution, or harmful prompt injection; the scripts perform standard file operations with basic safety checks against path traversal.
Capability Tags
Capability Assessment
Purpose & Capability
Name/description match the included artifacts: documentation, templates, scripts to create .learnings/, and OpenClaw hook handlers for injecting a reminder. The files and hooks are proportionate to a support-learning capture workflow.
Instruction Scope
Runtime instructions direct the agent to create and append to local .learnings/ files in the project or OpenClaw workspace and to use hooks for optional automation. The skill warns not to log PII but cannot enforce anonymization; enabling the PostToolUse hook runs a script that reads the CLAUDE_TOOL_OUTPUT environment variable (provided by the host) and emits detection reminders. This behavior is within scope but worth noting because it touches command output and local files.
Install Mechanism
No install spec or external downloads are required. The package includes local scripts and hook handlers only. Manual install instructions point to a GitHub repo and copying hooks into the OpenClaw hooks directory, which is a standard, low-risk pattern.
Credentials
The skill declares no required credentials or config paths. Scripts read CLAUDE_TOOL_OUTPUT and rely on OpenClaw/Claude hook environments, which is normal for detection hooks but that environment var is not listed as a required env in the manifest—this is expected for hook-driven scripts but worth calling out. No secrets or unrelated cloud credentials are requested.
Persistence & Privilege
always is false and the skill is user-invocable. Hooks inject a virtual reminder file at agent bootstrap and scripts can create files under the workspace or ./skills (extract-skill.sh), but nothing forces persistent global changes or modifies other skills' configs. Enabling hooks is an explicit opt-in step.
How to Use
- Make sure OpenClaw is installed (local or Docker)
- Run the install command in chat:
/install self-improving-support - After installation, invoke the skill by name or use
/self-improving-support - Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.1.0
**Version 1.1.0**
- Added stackability contract for multi-skill installations.
- Added namespaced logging guidance (`.learnings/support/`) for coexistence with other skills.
- Added required `Skill: support` metadata field and cross-skill precedence/ownership rules.
- Clarified hook arbitration model (single dispatcher, dedupe, rate limiting).
v1.0.0
self-improving-support v1.0.0
- Initial release enabling continuous support improvement through structured capture of ticket issues, delays, SLA breaches, misdiagnoses, knowledge gaps, escalation gaps, and churn signals.
- Introduces markdown log files: LEARNINGS.md, TICKET_ISSUES.md, and FEATURE_REQUESTS.md under a dedicated `.learnings/` directory.
- Provides clear setup instructions for OpenClaw and generic agent environments.
- Defines actionable logging formats and guidance for the promotion of high-impact learnings to KB articles, escalation matrices, and canned responses.
- Emphasizes anonymization and exclusion of customer PII or credentials in all logs.
Metadata
Frequently Asked Questions
What is Self-Improving Support?
Captures ticket resolution delays, misdiagnoses, escalation gaps, SLA breaches, knowledge gaps, and customer churn signals to enable continuous support impro... It is an AI Agent Skill for Claude Code / OpenClaw, with 94 downloads so far.
How do I install Self-Improving Support?
Run "/install self-improving-support" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.
Is Self-Improving Support free?
Yes, Self-Improving Support is completely free, licensed under MIT-0. You can download, install and use it at no cost.
Which platforms does Self-Improving Support support?
Self-Improving Support is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).
Who created Self-Improving Support?
It is built and maintained by José I. O. (@jose-compu); the current version is v1.1.0.
More Skills