localhero

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (high risk: 0.70). The skill explicitly loads project glossary and style settings from the external Localhero.ai service at runtime (see SKILL.md and README references to running npx @localheroai/cli glossary and npx @localheroai/cli settings), which are user-managed third-party content that the agent is instructed to read and use to guide its writing/decisions, creating a clear pathway for indirect prompt injection.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.90). The skill runs the CLI at activation (e.g., npx @localheroai/cli glossary/settings), which fetches/executes remote code via the npm package (https://www.npmjs.com/package/@localheroai/cli) and pulls glossary/settings from Localhero.ai (https://localhero.ai) that are then injected into and directly control agent prompts/instructions.