Skills

The skill permits arbitrary command injection and

The skill allows arbitrary `firecrawl`

This skill grants broad Bash execution permissions, enabling arbitrary command

The skill misrepresents itself as an active Firebase security rules

The skill deceptively claims to deploy applications but only

The skill allows raw SQL string literals, enabling SQL injection and data exfiltration, posing a significant security risk.

No security issues detected in expo/skills/expo-ui-swift-ui.

No security issues detected in firebase/agent-skills/firebase-ai-logic-basics.

The skill introduces supply chain risks via `curl | bash` installation and prompt injection vulnerabilities by directly interpolating user input into AI prompts.

The skill exposes native functions to untrusted web content

No security issues detected in expo/skills/expo-module.

No security issues detected in expo/skills/expo-ui-jetpack-compose.

The skill is vulnerable to prompt injection, allowing sub-

No security issues detected in Fleron/Claude-plugins/create-feature.

No security issues detected in Fleron/Claude-plugins/brainstorm.

No security issues detected in anthropics/claude-code/skill-development.

No security issues detected in anthropics/claude-code/mcp-integration.

This skill describes a plugin architecture

The skill allows command injection and persistent malicious execution via user

The skill grants broad file system access and inject

This skill enables severe command injection and arbitrary file system access

The skill misrepresents its capabilities,

This agent skill allows arbitrary command execution,

No security issues detected in anthropics/claude-code/writing-rules.

The skill is vulnerable to prompt and command injection, allowing

The skill uses dogmatic language to rigidly enforce

This skill enables arbitrary code execution and system compromise by executing user-defined plans from arbitrary file paths.

The skill risks resource exhaustion from

The skill delegates execution to sub-skills, expanding the attack surface and risking exploitation of downstream vulnerabilities or control bypass.

Designed to write a shared document, this skill poses a

This skill allows subagents to

The skill uses a hardcoded phrase as

This skill executes powerful cloud commands and is vulnerable to

The skill misrepresents its capabilities,

The skill uses prompt injection to override agent autonomy, enforce

This Azure diagnostics skill is vulnerable to command injection via

The skill's description indicates sensitive operations like password resets and network troubleshooting that could be abused if not properly secured.

The skill attempts to jailbreak the agent,

No security issues detected in microsoft/azure-skills/entra-app-registration.

The skill exposes direct Azure CLI

The Azure Resource Lookup skill allows arbitrary shell command injection through unsanitized user intent

The skill permits arbitrary Azure CLI command execution through injection due to insufficient input sanitization, despite a semantic read-

Exposes a tool for KQL queries on diagnostic logs, risking data exfiltration and reconnaissance if the agent is

No security issues detected in microsoft/azure-skills/azure-rbac.

No security issues detected in microsoft/azure-skills/azure-hosted-copilot-sdk.

The skill risks command injection by executing external tools with potentially unsanitized user input.

The skill allows arbitrary command execution and broad

The skill permits arbitrary Azure API calls