Skillv1.0.1

ClawScan security

Doc Format Converter · ClawHub's context-aware review of the artifact, metadata, and declared behavior.

Scanner verdict

SuspiciousApr 24, 2026, 10:41 AM

Verdict: suspicious
Confidence: medium
Model: gpt-5-mini
Summary: The skill largely matches a document-conversion tool, but there are multiple inconsistencies around billing/credentials, declared requirements, and SKILL.md vs. code that merit review before installing.
Guidance: Key issues to resolve before installing: - Metadata mismatch: the registry claims no env vars but SKILL.md and billing.py require SKILL_BILLING_API_KEY, SKILL_BILLING_SKILL_ID and FEISHU_USER_ID. Ask the publisher to correct the registry metadata so you know what secrets will be used. - Billing endpoint: the skill will POST user_id and charge data to https://skillpay.me. Confirm you trust that endpoint and understand that FEISHU_USER_ID will be transmitted for billing. - Instruction vs code mismatch: SKILL.md references functions (check_quota_free) and Feishu integration that are not present or not obviously implemented. Ask for a clear mapping of how billing/quota checks work and for the Feishu push code if the feature is advertised. - Undeclared system dependencies: OCR (pytesseract) and pandoc often require native binaries. Confirm whether the runtime provides tesseract/pandoc or if they will be installed, and whether those installs are safe. - Billing behavior: billing.py treats API-key-absent or API failures as a developer-mode success (returns large balance). This is a design choice but may mask network failures and could lead to unexpected free usage; request clarification. If you decide to proceed, only provide billing API keys and FEISHU_USER_ID if you trust the publisher and runtime, and consider running the skill in an isolated environment (sandbox) while you validate its behavior.

Review Dimensions

Purpose & Capability: noteThe code and SKILL.md implement a file conversion engine (pandas, Pillow, PyMuPDF, python-docx, pandoc fallbacks) which is consistent with the stated purpose. However the registry metadata lists no required environment variables or binaries while the SKILL.md and included billing.py clearly require SKILL_BILLING_API_KEY, SKILL_BILLING_SKILL_ID and FEISHU_USER_ID and the requirements.txt implies native tools (e.g., tesseract/pandoc) may be needed. This mismatch between declared metadata and actual needs is concerning.
Instruction Scope: concernSKILL.md and billing.py instruct the runtime to transmit FEISHU_USER_ID and billing requests to a third-party endpoint (https://skillpay.me/api/v1/billing). The conversion code itself operates on local files only, but SKILL.md also references Feishu notifications (no Feishu integration code present in the bundle). A sample in SKILL.md calls check_quota_free, but the provided billing.py does not define that function, indicating instructions and code are not fully aligned.
Install Mechanism: okNo install spec (instruction-only install) and no remote downloads; files are bundled in the skill. requirements.txt lists several Python packages which are expected for the functionality. No high-risk external URL downloads are present.
Credentials: concernThe skill uses and documents three sensitive environment values (SKILL_BILLING_API_KEY, SKILL_BILLING_SKILL_ID, FEISHU_USER_ID) but the registry metadata lists none. The FEISHU_USER_ID is sent to an external billing API — reasonable for billing but should be declared explicitly. Also requirements imply OCR (pytesseract) and pandoc which typically need system binaries; those system dependencies are not declared.
Persistence & Privilege: okThe skill does not request always:true, and it does not modify other skills or system settings. It will, however, perform network calls to the billing service during execution if API key is present.