The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes Evernote Note Markup Language (ENML) content through its validation and repair utilities, creating a surface for indirect prompt injection if malicious instructions are embedded in the processed notes.
Ingestion points: External note content is ingested via ENMLValidator.validate(content) and ENMLValidator.fix(content) in references/implementation-guide.md.
Boundary markers: The implementation lacks explicit delimiters or instructions for the agent to ignore embedded natural language commands within the note content.
Capability inventory: The skill has access to Bash(node:*) and Bash(npm:*) via the YAML frontmatter, and performs file system operations including fs.appendFileSync, fs.mkdirSync, and fs.renameSync.
Sanitization: ENMLValidator.fix implements regex-based sanitization to remove forbidden tags such as <script>, <form>, <style>, and <iframe>, and strips potentially dangerous attributes like onclick and onload.
[DYNAMIC_CONTEXT_INJECTION]: The skill uses dynamic context injection placeholders in the SKILL.md file to verify the local environment during skill loading.
Evidence: !node --version 2>/dev/null || echo 'N/A' and !python3 --version 2>/dev/null || echo 'N/A'.
Context: These commands are benign environment checks and do not involve network exfiltration or sensitive file access.

evernote-debug-bundle