The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill's setup instructions direct the user or agent to clone a repository from an untrusted personal GitHub account (github.com/m14r41/PentestingChecklist.git) and execute npm install. This involves downloading and potentially running lifecycle scripts from an unverified source.
[INDIRECT_PROMPT_INJECTION]: The skill implements features for importing data from external JSON strings (importFromJSON) and managing user-provided notes, which creates a surface for indirect prompt injection attacks.
Ingestion points: The importFromJSON function in SKILL.md processes external JSON data, and updateCheckNote ingests text content into the assessment state.
Boundary markers: There are no boundary markers or instructions to ignore embedded commands within the provided code snippets to prevent the agent from following instructions hidden in the imported data.
Capability inventory: The implementation snippets focus on client-side state management using localStorage and React; no high-risk capabilities like network exfiltration or shell execution are directly triggered by the data processing logic in the snippets.
Sanitization: The importFromJSON function performs a basic version compatibility check but lacks filtering or sanitization for natural language content within the checklist fields.

pentesting-checklist-interactive