Mental Atlas

PassAudited by VirusTotal on May 12, 2026.

Overview

Type: OpenClaw Skill Name: mental-atlas Version: 1.0.0 The skill bundle 'mental-atlas' is a structured knowledge extraction tool designed to distill complex information into a four-layer framework (Representations, Schemas, Mental Models, and Explanatory Frameworks). The SKILL.md file defines clear, non-malicious logic for processing text, files, and URLs using standard tools (Read and WebFetch) to generate educational summaries and test questions. There are no indicators of data exfiltration, prompt injection attacks, or unauthorized system access.

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.

What this means

If you ask it to process a local file or URL, the agent may load that material into the conversation for analysis.

Why it was flagged

The skill can read local files and fetch web content, which is expected for processing file paths and URLs but is still a capability users should understand.

Skill content
allowed-tools:
  - Read
  - WebFetch
Recommendation

Only provide files or URLs you are comfortable having analyzed by the agent.

What this means

A user-directed file path may cause the agent to read that file for summarization or distillation.

Why it was flagged

The skill instructs the agent to load user-provided local file paths. This is purpose-aligned for distillation, and the artifact does not show automatic broad file scanning.

Skill content
File path (starts with `/` or `~`, or ends with `.md`/`.txt`/`.pdf`) | Use Read tool to load, then process
Recommendation

Check that the requested path points only to the intended document and not to private or unrelated files.