The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests arbitrary background materials from users, such as old resumes, job descriptions (JD), and work notes, to extract information. This represents an indirect prompt injection surface where malicious instructions could be embedded in the provided documents.
Ingestion points: User-supplied background materials (SKILL.md, Section 1).
Boundary markers: None explicitly defined to instruct the agent to ignore instructions embedded within the user data.
Capability inventory: File writing, shell command execution (npm, python), and file copy operations.
Sanitization: No specific sanitization or validation of the input text is mentioned.
[DYNAMIC_EXECUTION]: The skill requires the agent to generate and execute JavaScript code using the docx library to produce the DOCX version of the resume.
Evidence: Code snippets for creating paragraphs and numbering are provided for the agent to implement at runtime (SKILL.md, Section 4).
[COMMAND_EXECUTION]: The skill instructs the agent to perform environment setup and output validation using shell commands.
Evidence: Instructions to run npm install -g docx for document generation and python scripts/office/validate.py for validating the generated DOCX file (SKILL.md, Section 4).

resume-builder