Observability & Reliability Engineering
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: afrexai-observability-engine Version: 1.0.0 The OpenClaw AgentSkills skill bundle 'afrexai-observability-engine' is classified as benign. The content, including the `SKILL.md` and `README.md` files, provides comprehensive documentation and examples for observability and reliability engineering. There is no evidence of intentional harmful behavior such as data exfiltration, malicious execution, persistence mechanisms, or obfuscation. The code snippets are illustrative, using standard libraries and tools (e.g., Pino, OpenTelemetry, Prometheus, kubectl) for their stated purpose. While some examples include powerful commands like `kubectl`, these are presented within the legitimate context of incident response runbooks and are not instructed for malicious use or self-exploitation by the agent. External links provided are for additional resources or related products, not for malicious downloads or command execution.
Findings (0)
Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.
If implemented carelessly, logs and dashboards could retain sensitive user or business data longer than intended.
The skill recommends structured observability logs that may contain persistent user and business identifiers, while also explicitly advising anonymization and secret/PII scrubbing.
client_ip: 203.0.113.42 # Anonymize in logs if needed ... business: user_id: "usr_456" ... amount_cents: 4999 ... Never log secrets ... PII & Secret Scrubbing
Use the provided scrubbing patterns, minimize logged fields, avoid secrets and direct PII, and enforce log retention and access-control policies.