Security audit

Cc Clean

Security checks for vulnerabilities and agentic risk

Overview

The skill appears intended to generate social content calendars, but its shipped artifact is incomplete and inconsistent about whether it runs locally or uses an Anthropic API key.

Review this skill before installing. It may require an Anthropic API key despite describing local generation, and the referenced generator script is missing from the artifact, so users cannot verify how prompts, outputs, API calls, or credentials would actually be handled.

A.I.G

Vulnerability Patterns

Insecure DependenciesIntroduces malicious components through unsafe dependency sources
Skill Instruction HijackingAlters the agent's session goals or safety constraints when the skill loads
Agent Memory PoisoningWrites attacker-controlled rules into memory that affect later sessions
Remote Payload Retrieval and ExecutionFetches external code whose behavior can change after review
Embedded Malicious CodeShips malicious scripts inside the skill and executes them locally

Findings (1)

T08 · Insecure Dependencies

Warning

Location: SKILL.md:33
Finding: Unpinned Third-Party Runtime Dependency
Content: View full analysis
Remediation: View remediation
bins: [] ``` 2. Maintain a lockfile that records all transitive dependency versions. 3. Require cryptographic hashes for downloaded distributions where supported. 4. Restrict dependency resolution to the official, trusted Python package index and disable untrusted supplemental indexes. 5. Run package installation and skill execution in a least-privilege sandbox with limited filesystem and network access. 6. Expose `ANTHROPIC_API_KEY` only to the process that requires it and prevent installation hooks from inheriting sensitive environment variables. 7. Add automated dependency scanning and review dependency updates before changing the pinned version. 8. Include the referenced `generate_calendar.py` implementation in the package so its dependency usage and credential handling can be audited. ]]>

SkillSpector

Vulnerability Patterns

MCP Tool PoisoningHidden Instructions, Unicode Deception, Parameter Description Injection
Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code

Findings (1)

Intent-Code Divergence

Medium

Category: Not specified by scanner
Confidence: 98% confidence
Finding: Line L178 says the skill uses 'Local MLX first (free, unlimited), Haiku fallback,' which describes a local model path as the primary behavior. However, the manifest metadata only declares an ANTHROPIC_API_KEY requirement and installation of the anthropic package, with no MLX dependency or setup described anywhere in the file. This is an active contradiction in the skill's own documentation about how generation is performed.
Content: No source excerpt is available for this finding.

Static analysis

No suspicious patterns detected.