Skillv1.1.0

ClawScan security

习惯养成教育追踪系统 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignApr 5, 2026, 8:47 AM

Verdict: Benign
Confidence: medium
Model: gpt-5-mini
Summary: The skill's code and instructions are consistent with a local family habit-tracking tool: it stores and queries a SQLite DB and offers CRUD/analysis scripts; there are some privacy/usability concerns (hard-coded DB path, no DB-init step, automatic recording rules) but no indications of exfiltration or unrelated privileges.
Guidance: This skill appears to be a local family habit-tracking tool and is internally consistent, but review these before installing: 1) Privacy: it stores personally sensitive data (children names, caregiver actions, behavioral notes) in a local SQLite DB — ensure the storage location is acceptable and has proper filesystem permissions. 2) Hard-coded DB path: the path is /home/istina/.openclaw/workspace/habit_education.db; update it if you run under a different user or want the DB elsewhere. 3) No DB-init: the bundle includes a schema.md but no script to create the DB/tables; create the DB with the provided schema before use or the scripts may fail. 4) Automatic recording rules: the skill auto-records 'father' participation — confirm this matches your privacy/recording policy. 5) Model context leakage: the agent will load stored records into context to decide duplicates and to answer queries — be mindful of what personal data you allow the agent to include in prompts. If you accept these caveats, run the skill in an environment you control, create the DB from references/schema.md, restrict file permissions, and test with non-sensitive data first.

Review Dimensions

Purpose & Capability: okName/description match the included scripts and schema: scripts implement new habit insertion, adding interventions, archiving, contributions, queries and stats against a local SQLite DB. The requested capabilities are proportional to a habit-tracking/analysis tool.
Instruction Scope: noteSKILL.md instructs the agent to load and use local DB state (active habits, records) as context and to automatically record certain actors (e.g., 'father' participation). That is expected for the feature but raises privacy/context-leak risk because stored child/family data may be surfaced into model prompts. The instructions otherwise do not reference unrelated files, network endpoints, or environment variables.
Install Mechanism: okNo install spec or remote downloads; the skill is delivered as code files and SKILL.md. There is no external package installation or URL-based download to review.
Credentials: noteThe skill requests no credentials or env vars (good). It hardcodes a DB path (/home/istina/.openclaw/workspace/habit_education.db), which is plausible but brittle and could cause unexpected cross-user access or data confusion if deployed in a different environment. The schema is provided but there is no included DB-init script; the skill assumes the DB with correct schema exists or will be created by the environment.
Persistence & Privilege: okSkill is not set to always: true, it does not request elevated privileges, and it does not modify other skills or system-wide settings. It persists data locally in its DB file only.