Skillv1.0.0

ClawScan security

阿德勒思维视角 · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

BenignApr 29, 2026, 2:25 PM

Verdict: benign
Confidence: high
Model: gpt-5-mini
Summary: This is an instruction-only roleplay skill that coherently implements an 'Adler perspective' persona and includes supporting research files; it requests no installs, credentials, or system access, so its footprint matches its stated purpose.
Guidance: This skill is an instruction-only roleplay that asks the agent to speak as Alfred Adler and bundles research notes — there are no downloads, installs, or requested secrets, so technical risk is low. Things to consider before installing: (1) It roleplays a deceased historical figure and uses first-person framing; do not treat its responses as actual historical quotations or as a substitute for licensed clinical care. (2) The SKILL.md’s single-disclaimer-once rule may lead to later replies appearing authoritative; if you plan to use this for sensitive contexts (mental health, crisis situations, clinical decision-making), require explicit, repeated disclaimers and human oversight. (3) Verify any clinical or factual claims against primary sources if you need high accuracy. If you want stronger safety controls, ask the developer to modify SKILL.md to (a) repeat the disclaimer before each session that includes clinical guidance, and (b) avoid definitive medical advice or include explicit routing instructions to professional help for crisis content.

Purpose & Capability: okName/description describe an Adler-perspective role and the SKILL.md plus reference files implement that goal. There are no unrelated env vars, binaries, or config paths requested.
Instruction Scope: noteInstructions require the agent to respond in-character as Alfred Adler ('I' voice) and to emit a single disclaimer only at first activation. This is coherent with the stated purpose but can be misleading to end users (strong persona/authority, single-use disclaimer). Consider requiring the persona to repeat or make the non-literal nature explicit in longer sessions or before giving clinical advice.
Install Mechanism: okNo install spec and no code files are executed — instruction-only skill. Low installation risk.
Credentials: okThe skill requests no environment variables, credentials, or config paths. The research/reference files are local to the skill bundle and appropriate for its educational/roleplay purpose.
Persistence & Privilege: okalways is false and there is no indication the skill modifies other skills or system settings. Model invocation is allowed (default) which is expected for a user-invocable roleplay skill.