Skillv1.0.2

ClawScan security

佛学导师 (The Buddhist Mentor ) · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

BenignApr 30, 2026, 10:19 AM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: The skill's files, instructions, and requirements are internally consistent with a guided Buddhist-mentor purpose; it is instruction-only, requests no external credentials or installs, and its behavior aligns with the stated goal.
Guidance: This skill is coherent and appears to do what it says: a structured, reflective Buddhist mentor. It is instruction-only, requires no external credentials, and makes no network calls. Two practical considerations before installing: (1) The skill's protocols instruct the agent to record and persist "important concepts" and "user personal turning points" into memory for future recall — if you do not want personal or sensitive information stored, disable or review agent persistent memory settings first. (2) The publisher/source has no homepage listed; if provenance matters to you, consider asking the publisher for contact or a canonical source before enabling. Otherwise the skill poses low technical risk.

Review Dimensions

Purpose & Capability: okName/description match the provided materials. All included files are pedagogical (glossary, guides, protocols). The skill requests no binaries, env vars, or external services that would be unrelated to a teaching mentor.
Instruction Scope: noteRuntime instructions are narrowly scoped to consulting the included files and following the described six-step interaction protocol (stage assessment, glossary lookups, constrained outputs, cross-disciplinary bridging, session reflection, memory management). No instructions tell the agent to read unrelated system files or call external endpoints. Note: the skill explicitly requires marking and persisting "important concepts" and "personal turning points" into the agent's memory system — this is coherent with a mentoring skill but has privacy implications (see guidance).
Install Mechanism: okInstruction-only skill with no install spec and no code files — nothing is written to disk and no third-party packages or downloads are requested.
Credentials: okNo environment variables, credentials, or system config paths are required. The only non-trivial resource is the agent's persistent memory functionality (referenced in Memory_Management.md), which is proportionate to a tutoring/mentoring use-case but should be considered for privacy.
Persistence & Privilege: noteThe skill does not request always:true nor modify other skills. It expects to use the agent's memory (marking important concepts and scheduled review triggers). This is within scope, but persistent storage of user statements (including sensitive personal experiences) is recommended to be reviewed by the user before enabling.