The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest structured data, drafts, and content from external sources (e.g., social media citations) and render it in a browser. It lacks explicit instructions to sanitize or escape HTML entities in the {content} placeholder, which creates a Cross-Site Scripting (XSS) vulnerability surface if the ingested data contains malicious scripts.
Ingestion points: Data identified from files, variables, or recent output (e.g., _private/drafts/outreach_drafts.md).
Boundary markers: Absent.
Capability inventory: File writing, file renaming (mv), and shell command execution (open).
Sanitization: Absent.
[DYNAMIC_EXECUTION]: The skill generates HTML files that include JavaScript logic for UI features like truncation toggles, clipboard copying, and local storage management at runtime.
Evidence: The assets/base-template.html and references/full-guide.md files provide script blocks that the agent is instructed to include in the generated output files.
[COMMAND_EXECUTION]: The skill uses the open command to launch the generated HTML files in the user's default web browser.
Evidence: SKILL.md contains the instruction: "Open with open _private/views/{filename}."

quick-view