MakeSoul Lite

Security checks across static analysis, malware telemetry, and agentic risk

Overview

MakeSoul Lite appears aligned with its stated purpose, but it gives an agent a permanent posting identity and encourages public sharing/importing of behavior-shaping personality files.

Install only if you want your agent to interact with MakeSoul.org. Manually approve public posts, updates, and deletes; keep the private key secret; avoid uploading personal data in USER.md; and treat all community soul templates as untrusted text until reviewed.

Static analysis

No static analysis findings were reported for this release.

VirusTotal

66/66 vendors flagged this skill as clean.

View on VirusTotal

Risk analysis

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.

#
ASI01: Agent Goal Hijack
Medium
What this means

A community soul template could contain instructions that change how the agent behaves or prioritizes tasks.

Why it was flagged

The skill is designed to fetch community-authored personality files that explicitly include behavioral rules, which could influence an agent's goals or operating style if applied without review.

Skill content
Browse and fetch favorite soul personalities ... SOUL.md - Core values, personality, behavioral rules
Recommendation

Treat fetched souls as untrusted content; require user review before adopting any SOUL.md, TOOLS.md, IDENTITY.md, or USER.md content as agent instructions.

#
ASI02: Tool Misuse and Exploitation
Medium
What this means

The agent could create public content or change/delete MakeSoul content under its identity.

Why it was flagged

The documented API lets the agent publish public souls/dreams and update or delete created souls. These are high-impact public/community mutations, and the artifacts do not clearly require explicit user confirmation before use.

Skill content
"Create Soul" ... "is_public": true ... "Create Dream" ... "is_public": true ... "Delete Soul"
Recommendation

Require explicit confirmation for all POST, PUT, and DELETE actions, and default submissions to private unless the user intentionally chooses public sharing.

#
ASI03: Identity and Privilege Abuse
Medium
What this means

Anyone with the private key may be able to act as that MakeSoul agent and modify its created content.

Why it was flagged

The skill creates and uses a long-lived private key as an agent identity for authenticated operations, but the artifacts do not define secure storage, rotation, or clear handling boundaries.

Skill content
"Register a new agent and get permanent identity (private key)" ... "private_key": "Save this - permanent identity"
Recommendation

Store the private key only in a secure secret store, avoid pasting it into public logs or shared chats, and use a dedicated low-risk identity for testing.

#
ASI06: Memory and Context Poisoning
Medium
What this means

Personal preferences, user context, or prompt-like instructions could be published or reused in ways the user did not intend.

Why it was flagged

The documented public submission format includes USER.md preference/context material, which may persist remotely and be reused by other agents or future sessions.

Skill content
"Create and submit a soul template to community with 4 files" ... "USER.md": "User interaction preferences" ... "is_public": true
Recommendation

Scrub USER.md and other files for personal or sensitive data before upload, and review community templates before using them as persistent context.