The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied Markdown files, which serves as a potential ingestion point for indirect instructions. However, the rendering core in scripts/slide_render.py implements security-best-practice sanitization by HTML-escaping all content before applying Markdown formatting and enforcing a strict allowlist for URLs (SAFE_URL_RE). This prevents untrusted Markdown from injecting malicious scripts or hijacking the presentation context.
[REMOTE_CODE_EXECUTION]: Analysis of the Python scripts confirms the absence of remote code execution patterns. The tools operate strictly as local command-line utilities, relying only on standard libraries and explicit local file inputs. There are no calls to eval(), exec(), or subprocesses that execute external code.
[DATA_EXFILTRATION]: No network operations or data exfiltration vectors were identified. The scripts lack any network-capable code (e.g., requests, urllib, or socket), and the generated HTML decks are designed to be self-contained for offline use.
[CREDENTIALS_SAFE]: A search of all files (scripts, templates, and styles) confirmed that no hardcoded API keys, tokens, or sensitive credentials are present.

md-slides