Triage
PassAudited by VirusTotal on May 12, 2026.
Overview
Type: OpenClaw Skill Name: triage Version: 1.0.0 The OpenClaw AgentSkills skill bundle 'triage' is benign. All files (`_meta.json`, `SKILL.md`, `patterns.md`, `signals.md`) contain instructions and documentation solely focused on teaching an AI agent to prioritize tasks based on urgency, impact, and user patterns. There is no evidence of malicious intent, such as data exfiltration, unauthorized command execution, persistence mechanisms, or prompt injection attempts designed to subvert the agent's core function or access sensitive data. The instructions are clear and directly support the stated purpose of task triage.
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.
Urgency wording or learned P0 patterns could cause the agent to reorder or interrupt other tasks.
The skill gives priority labels control over execution order, so an incoming item classified as P0 can interrupt current work. This is disclosed and central to triage, but it affects agent behavior.
Route — P0 immediately, P1-P3 into queue ... New P0 interrupts P1 work
Treat automatic P0 routing as advisory for important work, and ask the agent to confirm interruptions when the impact is unclear.
Incorrect or overly specific learned patterns could cause future tasks to be prioritized poorly, and recorded context may preserve sensitive task hints.
The skill asks the agent to retain override context and learned priority patterns for future decisions. This persistent context could include task details and influence later triage.
When user explicitly overrides: 1. Record the override with context ... # Recent overrides (last 10)
Review learned patterns periodically, avoid storing sensitive details in priority notes, and require confirmation before applying permanent or locked rules.