Evomap Skills

What This Skill Does

Evomap Skills is a high-productivity utility pack curated by 2233admin, designed to bridge critical gaps in automated agent performance and infrastructure reliability. This comprehensive toolkit provides eight specialized modules categorized into code-based logic and knowledge-based frameworks. At its core, the skill addresses the common pain points of AI agents, such as connection instability, formatting failures in communication channels like Feishu, and the complexities of managing memory and performance within orchestrated environments like Kubernetes. By integrating this toolkit, developers can significantly reduce the overhead of manual troubleshooting and enable their agents to self-correct during common runtime failures.

Installation

To integrate these utilities into your OpenClaw environment, execute the following command in your terminal:

clawhub install openclaw/skills/skills/2233admin/evomap-skills

Ensure that your OpenClaw instance has appropriate write permissions to the plugin directory and that you have authenticated with your credentials to access the central repository.

Use Cases

• Service Reliability: Utilize http-retry to handle intermittent network failures, ensuring that your data fetches are robust against 429 rate limiting and connection resets.
• Communication Resilience: Implement feishu-fallback to ensure critical alerts are delivered regardless of rendering errors, automatically shifting from complex card formats to simplified text versions.
• Infrastructure Optimization: Use k8s-memory-tune to automatically manage JVM heap settings based on OOM (Out-of-Memory) signals, preventing container crashes in resource-constrained environments.
• Complex Task Coordination: Deploy swarm-task to delegate sub-tasks to multiple agents, enabling efficient parallel processing of large-scale requests.
• Self-Healing Agents: Leverage agent-self-debug to trigger automated diagnostic loops when an agent encounters unknown exceptions, reducing downtime.

Example Prompts

1. "The agent keeps failing to deliver alerts in the Feishu group, please apply the feishu-fallback fix to ensure they reach the team."
2. "My service is being OOM killed in Kubernetes; use the k8s-memory-tune skill to analyze and optimize the JVM heap limits."
3. "We are seeing high latency on external API calls, please configure the http-retry mechanism to handle 429 requests and connection resets."

Tips & Limitations

• Configuration: Always test the retry logic and fallback mechanisms in a staging environment before deploying to production to avoid flooding channels with duplicate errors.
• Limitations: The memory-tuning tools require proper RBAC permissions to interact with your Kubernetes cluster safely. Ensure that your agent service account has the necessary read/write scope for namespace metrics.
• Monitoring: While metric-anomaly provides insights into performance, it is intended to complement, not replace, formal APM (Application Performance Monitoring) systems.