Chinese Tutor

v1.0.0

Proactive Chinese language tutor that delivers curated, real-world Mandarin learning content on a schedule. Use when: (1) User wants to learn or improve Chin...

2· 214·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
VirusTotalVirusTotal
Benign
View report →
OpenClawOpenClaw
Benign
medium confidence
Purpose & Capability
The skill's name/description (proactive Chinese tutor) matches the instructions: curated lessons, content sourcing, OCR for user documents, and maintaining a learner profile. One notable omission: it describes scheduled 'pushes' and preferred_times but does not declare or request any notification/delivery credentials or explain the platform mechanism for scheduling/delivery (cron/heartbeat is referenced conceptually). This is coherent if the host platform provides scheduling/notification features, but the SKILL.md assumes such capabilities without documenting how to connect to email/messaging/push channels.
Instruction Scope
SKILL.md instructs the agent to create and persist a learner profile (hinihao-profile.json), append vocab/grammar notes, and generate lesson files; it also directs OCR of user-uploaded PDFs/DOCX/images via native vision and to extract and save vocabulary/grammar. All of this is consistent with a tutor skill, but it means the skill will read user-uploaded documents and write persistent files — a privacy consideration. There are no instructions to transmit sensitive files to external endpoints or to access unrelated system paths.
Install Mechanism
No install spec and no code files — instruction-only. This minimizes disk footprint and risk from arbitrary installers. Nothing is downloaded or executed beyond the agent following prose instructions.
Credentials
The skill requires no environment variables, credentials, or config paths. That is proportionate to its stated purpose. The only persistent artifacts described are local profile and lesson files; there are no requests for unrelated secrets or external API keys.
Persistence & Privilege
The skill persists learner data (profile JSON, vocab_bank, grammar notes, lesson markdown files) and updates them over time. always:false (not force-included) and default model invocation are normal. The write/persistence behavior is expected for a tutor but is a capability the user should be aware of — especially for uploaded documents processed by OCR.
Assessment
This skill appears internally consistent with its tutoring purpose and doesn't ask for credentials or install anything, but review these items before installing: - Delivery/scheduling: The skill talks about scheduled 'pushes' (preferred_time, push_times) but doesn't specify how messages are delivered (email, SMS, platform notifications, external webhook). Confirm with the platform how scheduled lessons are delivered and whether additional connectors/permissions are needed. - Local persistence: The skill will create and update files (e.g., hinihao-profile.json, vocab/grammar notes, lesson markdown). Decide if you accept those files being stored where the agent runs and whether you need to back them up or encrypt them. - Uploaded content / privacy: Document Study and Snap & Learn use OCR on user-uploaded PDFs/images and will extract text and save vocabulary/notes. Do not upload sensitive documents (IDs, private contracts, personal medical records) unless you trust the runtime storage and retention policy. - External links and accessibility: The skill recommends content from Chinese platforms (Bilibili, Douyin, WeChat, etc.). Some links or content may be region-restricted; the agent may provide search terms rather than direct links. Expect the agent to perform web lookups to verify creators and freshness. - If you need stronger guarantees: ask the skill author (or the platform) how long persisted files are retained, where they are stored, whether they are shared outside your account, and what mechanism is used for scheduling/notifications. Overall: coherent and plausible for its purpose, but treat persisted learner data and uploaded documents with appropriate caution.

Like a lobster shell, security has layers — review code before you run it.

latestvk97d2hmk3bxe7p942ay0kh9hes82rhbm

License

MIT-0
Free to use, modify, and redistribute. No attribution required.

Comments