Skillv1.8.0

ClawScan security

Isfp Coach · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

ReviewApr 17, 2026, 10:41 AM

Verdict: Review
Confidence: medium
Model: gpt-5-mini
Summary: The skill's functionality (an ISFP coaching assistant) is coherent and it requests no credentials, but the package contains inconsistent/mislabeled files and unclear data‑storage instructions, which warrant caution before installing.
Guidance: This skill appears to be a legitimate ISFP coaching assistant and requests no credentials, but several red flags suggest sloppy packaging and unclear privacy behavior. Before installing: 1) Ask the publisher to fix metadata (README, clawhub.json, repository URL) and confirm the skill is indeed authored for ISFP. 2) Ask where '成长记录' (user archives) are stored: confirm they remain in-agent/local only, or if sent to a remote server, request the endpoint, purpose, and data retention policy. 3) Verify explicit user consent and deletion procedures for stored conversations. 4) If possible, test the skill in a sandboxed agent with no real PII to confirm it only stores data locally. 5) If you need stronger assurance, request a corrected repo or source link and an updated manifest that matches the SKILL.md. If these questions are unanswered or the package metadata remains inconsistent, consider not installing or treat the skill as untrusted for sensitive data.

Review Dimensions

Purpose & Capability: noteThe SKILL.md describes an ISFP coach and the runtime instructions map to that purpose (coach vs. advisor modes, GROW model, brief replies, etc.). However the package contains multiple mismatches: README.md is an INTP coach, clawhub.json keywords and repository URL point elsewhere, and some reference documents are labeled/located under 'intp-coach'. These look like copy/paste or packaging errors and reduce confidence that the published metadata accurately represents the skill.
Instruction Scope: concernThe instructions explicitly describe recording and archiving '用户档案' (user's key insights, actions, and progress) and repeatedly say to 'store' conversation and action records. The skill does not document where or how these archives are stored (local agent storage, remote endpoint, etc.). There is no mention of consent, retention, or deletion. While recording is plausible for a coach skill, the lack of storage detail is a privacy concern and should be clarified.
Install Mechanism: okThis is an instruction-only skill with no install spec and no code files to execute; nothing is downloaded or written during install. That minimizes installation risk.
Credentials: okThe skill declares no required environment variables, no credentials, and no special config paths. Requested capabilities are proportionate to an instruction-only coaching skill.
Persistence & Privilege: noteThe skill is not always-enabled and does not request elevated privileges. However it instructs the agent to 'save' and 'archive' user records across sessions—this implies persistence. Because the package doesn't specify storage scope (agent-local vs remote) or retention policies, treat its implied persistence as a privacy consideration rather than a platform privilege escalation.