Lumi Diary

ReviewAudited by ClawScan on May 10, 2026.

Overview

Prompt-injection indicators were detected in the submitted artifacts (unicode-control-chars); human review is required before treating this skill as clean.

Install Lumi Diary if you want a local memory companion and are comfortable with it saving diary fragments, media references, contacts, traits, and milestones in a local Lumi_Vault folder. Use it only in groups where people know it is present, be careful importing shared capsules, and keep backups if you may edit or delete memories. ClawScan detected prompt-injection indicators (unicode-control-chars), so this skill requires review even though the model response was benign.

Findings (5)

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

Private facts, personality notes, birthdays, anniversaries, and group-member details may persist locally and shape later responses.

Why it was flagged

The skill intentionally records personal and social profile information and reuses it later. This is central to a memory diary, but it creates sensitive persistent context.

Skill content
Call `update_portrait` to record personality traits, preferences, and milestones — then weave them into future interactions.
Recommendation

Use it only in contexts where recording is expected, inform group members, and periodically review or back up/delete the Lumi_Vault data if needed.

What this means

A mistaken update or delete request could alter or remove local memory records.

Why it was flagged

The skill can modify and delete stored diary fragments. This is expected for a diary manager, but it is still local data mutation authority.

Skill content
`manage_fragment` | Full CRUD: search/get/update/delete fragments.
Recommendation

Confirm destructive requests, keep backups of the vault if memories matter, and avoid granting the skill access to irreplaceable data without a recovery plan.

What this means

A capsule from someone else could add inaccurate, unwanted, or manipulative memories to the local diary.

Why it was flagged

The import feature can persist externally supplied memory content and media in the local vault. This is user-directed and purpose-aligned, but imported content may later appear in canvases or context.

Skill content
Friend: "Lumi, import this capsule" ... merges annotations, copies media into local vault
Recommendation

Import capsules only from trusted people, inspect shared files when possible, and remove unwanted imported entries promptly.

What this means

Any trusted MCP client you configure may be able to search, write, export, or import Lumi diary data.

Why it was flagged

The MCP adapter exposes the diary tools to configured agent clients. This is documented integration behavior, but those clients can interact with the local memory vault through the server.

Skill content
Wraps ``lumi_core`` functions as MCP tools for any MCP-compatible client (Claude Desktop, Cursor, VS Code Copilot, etc.).
Recommendation

Enable the MCP server only for clients you trust and avoid running it in shared or untrusted environments.

What this means

Users have less external information to verify the publisher or compare the installed package against an upstream repository.

Why it was flagged

The registry metadata does not provide a public source or homepage for provenance checks. The provided artifacts do not show hidden installers or remote scripts, so this is a provenance note rather than a concern.

Skill content
Source: unknown; Homepage: none
Recommendation

Install from the trusted registry entry, review the local package contents if high assurance is needed, and be cautious with optional dependencies.