i-skill

PassAudited by ClawScan on May 1, 2026.

Overview

The skill is coherent and purpose-aligned, but enabling it creates a persistent personalization profile from conversations that can shape later assistant responses.

This appears safe to install if you want persistent personalization. Before enabling it, understand that it will create local profile files from conversation context and use them to tailor later responses; review or delete the profile if it becomes inaccurate or too sensitive.

Findings (3)

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

After enabling personalization, the assistant may keep applying stored assumptions about the user until the profile is paused, updated, reset, or deleted.

Why it was flagged

The skill persists a conversation-derived user profile and reuses it as assistant context after activation, which is central to the purpose but means personal preferences and any inaccurate profile content may influence later responses.

Skill content
The profile is stored in `myself.md` ... If `status == "active"`, read `myself.md` ... and use it as personalization context
Recommendation

Use the view, update, pause, reset, and delete commands as needed, and avoid storing sensitive details you do not want reused for personalization.

What this means

Once activated, responses may be tailored according to the stored profile even when the user does not explicitly ask for personalization in each message.

Why it was flagged

The skill deliberately changes assistant behavior across the active session. This is disclosed and purpose-aligned, but it is a broad instruction scope.

Skill content
Behavioral Anchors (High Priority — Always Obey) ... remain active throughout the entire session after activation ... every response should reflect understanding of the user profile
Recommendation

Activate it only when you want session-wide personalization, and pause it when you want neutral responses.

What this means

Users have less external context for verifying the publisher or project history before installing.

Why it was flagged

The registry metadata does not provide an upstream source or homepage, which limits provenance context for a skill that handles personal profile data.

Skill content
Source: unknown; Homepage: none
Recommendation

Review the included artifacts and install only if you trust the listed owner and are comfortable with local personalization storage.