The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx to fetch and execute the @chasen-liao/resume-skills package from the npm registry at runtime. This package is a vendor resource owned by the skill author. The @latest tag is specified to ensure the most recent version is downloaded for validation and editing tasks.
[COMMAND_EXECUTION]: Invokes shell commands through npx for validating resume structures (resume-skills validate) and launching a local editor service (resume-skills editor). It also executes a PowerShell script (render_resume.ps1) to handle the document rendering process and metadata generation.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted job description (JD) content, which could potentially contain malicious instructions intended to influence the agent's behavior.
Ingestion points: Step 1 and 2 in SKILL.md where the agent reads user-provided JD text or uploaded files.
Boundary markers: The skill instructs the agent to strictly follow established facts in resume-facts.yaml and adhere to the guidelines in resume-contract.md, which serves as a logical constraint against hallucination or unauthorized modification.
Capability inventory: Capability to write customized HTML, PDF, and Markdown files to the local file system and execute validation/rendering commands.
Sanitization: No explicit sanitization or filtering of the job description input is described in the workflow.
[DYNAMIC_EXECUTION]: The skill utilizes a local web server for live previewing of resumes and executes a local PowerShell script to dynamically generate PDF files and manifest metadata during the tailoring process.

jd-tailorer