ARG Action Chain Designer
PassAudited by VirusTotal on May 8, 2026.
Overview
Type: OpenClaw Skill Name: arg-action-chain-designer Version: 1.0.0 The skill bundle is a design tool for creating structured, multi-step agent workflows called 'ARG Action Chains.' The instructions in SKILL.md focus on engineering reliability, implementing validation gates, and reducing agent hallucination through deterministic scripts and bounded judgment. No indicators of data exfiltration, malicious execution, or harmful prompt injection were found; the content is entirely aligned with its stated purpose of workflow optimization.
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.
A generated action chain may include shell commands or scripts that a user or agent later executes.
The skill can design workflows that include commands or scripts. This is central to its purpose, but those generated commands could affect files or systems if later run without review.
Validation gates: commands, schema checks, checklist gates, or human review gates.
Review generated commands and scripts before running them, especially if they modify files, call external services, or perform irreversible actions.
Users could accidentally include real secrets or overbroad access details in a generated workflow design.
The skill may ask the user to account for credentials or external-system access when designing a workflow. The artifacts do not show credential collection or transmission, but this is sensitive information to handle carefully.
Inputs: files, APIs, user context, credentials, schedules, or external systems.
Describe credential requirements by type and scope rather than pasting secret values; use least-privilege credentials and approved secret storage for any workflow that later runs.