The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands to manage files and execute auxiliary scripts.
Evidence: Executes mkdir -p, open, bash, and python to perform deck generation, PDF export, and PowerPoint extraction.
[EXTERNAL_DOWNLOADS]: The skill initiates external downloads to set up its execution environment.
Evidence: The export-pdf.sh script automatically installs the Playwright browser automation library when first run.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a vulnerability surface for indirect prompt injection by processing untrusted data.
Ingestion points: Processes user-provided notes, URLs, and uploaded PowerPoint files (via extract-pptx.py) in SKILL.md Step 1 and Step 5.
Boundary markers: No explicit delimiters or instructions are used to prevent the agent from following commands embedded in the processed content.
Capability inventory: The agent has access to shell execution (bash, python), file system operations (mkdir, open), and network deployment (Vercel).
Sanitization: No sanitization or validation of the ingested content is specified before it is used to draft slide copy or determine deck structure.
[REMOTE_CODE_EXECUTION]: The skill depends on and executes code from an external directory on the host machine.
Evidence: It hardcodes paths to a sibling directory /Users/ksd/Desktop/Varnan_skills/frontend-slides/ to execute shell and Python scripts for core functionality like PDF export, deployment, and file extraction.

graphic-slide-deck