Tutor

• Risk level: Low
• Analyzed: May 1, 2026, 5:39 AM
• Findings: 2
• Version: 1.0.0

Overview

The skill is a coherent instruction-only tutor with no code or credentials, but it intentionally keeps local learner records and reports that may contain sensitive student information.

This appears safe to install as an instruction-only tutoring skill, but be comfortable with it creating local records for each learner. For children or students, obtain appropriate consent, keep the stored folder private, review any parent-facing reports or safety escalations before sharing, and delete records when no longer needed.

Findings (2)

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.

ASI06: Memory and Context Poisoning

Note

What this means

Learner profiles, session notes, progress data, and reports may remain on disk and influence future tutoring sessions.

Why it was flagged

The skill creates persistent local memory about learners and reloads it across sessions. This supports personalized tutoring, but the records can include sensitive information about minors or students.

Skill content

~/tutor/ ... profile.md # Age, grade, learning style, goals ... sessions.jsonl ... progress.json ... reports/ ... ALWAYS log sessions to ~/tutor/{learner}/

Recommendation

Use the skill with appropriate consent, keep the ~/tutor folder private, review records periodically, and delete learner data when it is no longer needed.

ASI08: Cascading Failures

Note

What this means

Sensitive comments from a learner may be summarized or raised to a guardian, which should be handled thoughtfully, especially for minors.

Why it was flagged

The skill can turn sensitive learner disclosures into parent/guardian escalation. This is aligned with the stated safety and parent-oversight purpose, but inaccurate or overly broad escalation could propagate sensitive information.

Skill content

Immediate Escalation to Parent/Guardian ... If learner mentions or implies: ... Abuse of any kind ... Flag immediately to parent/guardian

Recommendation

Have a responsible human review reports and safety escalations before sharing them, and consider the learner’s context and applicable safeguarding practices.