Security Audit

anysiteio/agent-skills:skills/anysite-cli

github.com/anysiteio/agent-skills

AI SkillCommit 34bedfab65ad

CAUTION

Scanned 12 days ago

Critical

Immediate action required

High

Priority fixes suggested

Medium

Best practices review

Low

Acknowledged / Tracked

Trust Assessment

anysiteio/agent-skills:skills/anysite-cli received a trust score of 65/100, placing it in the Caution category. This skill has some security considerations that users should review before deployment.

SkillShield's automated analysis identified 4 findings: 0 critical, 2 high, 2 medium, and 0 low severity. Key findings include Insecure Credential Handling in CLI Commands, SQL Injection Vulnerability via CLI Interface, Potential Data Exfiltration via Webhooks.

The analysis covered 4 layers: llm_behavioral_safety, manifest_analysis, static_code_analysis, dependency_graph. The llm_behavioral_safety layer scored lowest at 56/100, indicating areas for improvement.

Last analyzed on February 8, 2026 (commit 34bedfab). SkillShield performs automated 4-layer security analysis on AI skills and MCP servers.

Layer Breakdown

Manifest Analysis

100%

Static Code Analysis

100%

Dependency Graph

100%

LLM Behavioral Safety

56%

Behavioral Risk Signals

Network Access

1 finding

Filesystem Write

1 finding

Shell Execution

2 findings

Dynamic Code

2 findings

Security Findings4

Severity	Finding	Layer	Location
HIGH	Insecure Credential Handling in CLI Commands The skill documentation explicitly instructs the agent to pass sensitive API keys and database passwords as plain text command-line arguments. This practice exposes secrets to process listings, shell history, and agent logs, making them vulnerable to leakage. Update documentation and usage patterns to pass secrets exclusively via environment variables (e.g., ANYSITE_API_KEY, PGPASS) or a secure secrets manager, rather than CLI arguments.	Unknown	SKILL.md:45
HIGH	SQL Injection Vulnerability via CLI Interface The skill exposes a direct SQL execution interface (`--sql`) to the agent. If the agent constructs SQL queries by concatenating unsanitized user input, it creates a direct vector for SQL injection attacks against the connected database (PostgreSQL or DuckDB). Restrict the agent's ability to execute arbitrary SQL. Provide a set of predefined, parameterized queries or safe data access methods instead of a raw SQL interface.	Unknown	SKILL.md:230
MEDIUM	Potential Data Exfiltration via Webhooks The dataset configuration supports defining webhooks (`type: webhook`) that receive collected data. An attacker could manipulate the agent (via prompt injection) into configuring a malicious webhook URL to exfiltrate sensitive data to an external server. Implement a whitelist of allowed webhook domains or require explicit human approval before the agent can configure or execute data exports to external URLs.	Unknown	SKILL.md:185
MEDIUM	Unrestricted File System Write Capability The skill allows the agent to define export paths for data files using templates (e.g., `path: ./output/{{source}}-{{date}}.csv`). Without explicit path sanitization or sandboxing documented, this could allow the agent to overwrite sensitive system files if the `source` variable is manipulated (Path Traversal). Enforce a strict sandbox directory for all file write operations and sanitize all dynamic path components to prevent directory traversal attacks.	Unknown	SKILL.md:166

Scan History

Embed Code

[![SkillShield](https://skillshield.io/api/v1/badge/69d9eab16c9e1faf.svg)](https://skillshield.io/report/69d9eab16c9e1faf)