Capability Pipeline OS

PassAudited by ClawScan on May 16, 2026.

Overview

This is an instruction-only workflow planning skill with no code, credentials, or install actions, but users should supervise any high-impact tasks it helps execute.

This skill appears safe to install as an instruction-only planning framework. Treat it as a way to organize tasks, not as permission for the agent to act autonomously on sensitive systems; ask for approval checkpoints before any important edits, submissions, messages, transactions, or deployments.

Findings (1)

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.

What this means

If a user applies this workflow to code changes, business decisions, legal, medical, financial, or external-account actions, the agent may treat some steps as automatically executable unless the user sets boundaries.

Why it was flagged

The skill can structure tasks into units that may be executed automatically or semi-automatically. This is purpose-aligned for a pipeline framework, but it could amplify mistakes if applied to high-impact actions without user confirmation.

Skill content
AI自治度 | ⬛全自动 / 🟨半自动(需人工确认) / ⬜辅助(人工主导) ... 按管线顺序执行:遵循编排模式,依次或并行执行单元
Recommendation

Use the skill for planning and structured execution, but require explicit approval before irreversible changes, external communications, transactions, deployments, or edits to important files.