Skill flagged — suspicious patterns detected
ClawHub Security flagged this skill as suspicious. Review the scan results before using.
MedCrypt: End-to-End Encryption for Medical Messaging
v1.0.0Encrypts medical data client-side with AES-256-GCM for secure messaging on Telegram/WhatsApp, ensuring HIPAA, GDPR, and LFPDPPP compliance.
⭐ 0· 65·0 current·0 all-time
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Suspicious
medium confidencePurpose & Capability
The description claims end-to-end encryption for messaging on Telegram/WhatsApp and regulatory compliance. The included medcrypt.py implements AES-256-GCM encryption, PBKDF2 key derivation, key rotation, an audit-log abstraction, and Shamir sharing — all consistent with client-side encryption. However, there is no code that integrates with Telegram or WhatsApp, no QR-code generation/parsing code in the package, and no code that implements the stated 'encrypted transport' or compliance controls. The skill appears to provide crypto primitives rather than a full messaging integration.
Instruction Scope
SKILL.md simply instructs: 'pip install cryptography' and 'python medcrypt.py' and describes protocol/operational rules (QR PBKDF2, wire format, key rotation). The instructions are vague and over-broad: they reference QR-code-based key exchange and sending via Telegram/WhatsApp but do not provide code or commands to perform those actions. The medcrypt.py file shown performs local crypto and an in-memory encrypted audit log, but the SKILL.md asks users to 'review these carefully' and lacks operational details (how to supply secrets, how to persist logs, CLI/IO). The runtime instructions therefore give the agent broad discretion without concrete, safe steps.
Install Mechanism
No install spec in the registry; SKILL.md recommends 'pip install cryptography' before running the included Python script. This is a low-risk, standard dependency installation from PyPI. There are no downloads from unknown URLs or archive extraction steps in the metadata.
Credentials
The skill declares no required environment variables, credentials, or config paths. The code excerpt does not reference environment secrets or external credentials. This is proportionate to the provided functionality (local crypto primitives).
Persistence & Privilege
The skill is not always-enabled and does not request elevated privileges. The included audit log implementation stores entries in-memory (self.entries) in the shown code; no evidence in the provided excerpt of writing system-wide configuration, other skill settings, or network exfiltration. That said, persistence and transport behavior depends on the rest of the (truncated) file and how users run the script.
What to consider before installing
What to consider before installing/using this skill:
- Functional expectations: The package supplies local encryption primitives and a wire format, but it does NOT include Telegram/WhatsApp integration, QR-code generation/parsing, or automated transport. If you expect an out-of-the-box messaging integration, this is not it.
- Incomplete / buggy implementation: The provided medcrypt.py excerpt contains issues you should resolve before trusting it in production (for example, use of datetime.UTC, which is not a standard attribute in Python's datetime module, and the file appears truncated around the share-recovery function). Test the script in an isolated environment first.
- Provenance and compliance claims: The source and homepage are unknown. Legal claims like 'HIPAA/GDPR compliant' cannot be verified from the code alone—compliance depends on deployment, operational controls, and organizational practices. Do not rely on this claim without legal/security review.
- Cryptography review: While primitives appear sensible (AES-GCM, PBKDF2, CSPRNG), cryptographic correctness and side-channel/security details matter (deterministic salt derivation from the secret, iteration count impact, correct zeroization semantics, correctness of the Shamir implementation). If you plan to use this for real patient data, have a qualified crypto/security engineer review the full code or prefer well-vetted libraries/protocols.
- No network exfiltration detected in the provided excerpt, but verify the rest of the file: ensure no hidden endpoints, logging to disk, or secret exfiltration is present before running with real data.
- Operational recommendations: run the script in an isolated environment, perform a code review (complete file), add explicit tests for key exchange and recovery, and do not use for protected health information in production until reviewed and integrated with a secure transport layer and audited operational procedures.Like a lobster shell, security has layers — review code before you run it.
latestvk975ge4738prmh7e9fk7t6y8rx83dr7g
License
MIT-0
Free to use, modify, and redistribute. No attribution required.
