The Agent Skills Directory

[DYNAMIC_EXECUTION]: The verification suite (verify_code.py, verify_blocks.py, and verify_challenges.py) compiles and executes code snippets extracted from markdown drafting files. This includes running Node.js scripts and compiled Rust binaries to validate that the examples taught in the courses are functional.
[COMMAND_EXECUTION]: Multiple tools use the subprocess module to execute shell commands and system utilities. render_visuals.py calls weasyprint, pdfinfo, pdftoppm, mutool, and cwebp to transform HTML specs into images. ci.py executes the other internal validation tools.
[EXTERNAL_DOWNLOADS]: The environment setup (verify/Dockerfile) fetches and executes toolchain installers from several remote sources, including Solana Anza (release.anza.xyz), Rustup (sh.rustup.rs), NodeSource (deb.nodesource.com), and Foundry (foundry.paradigm.xyz). These are well-known and reputable sources in the development ecosystem.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process lesson drafts and user notes to generate content. Because the verification tools execute code extracted from these files, there is a risk that malicious instructions or code blocks embedded in a draft could be executed during the automated gate checks.
[DYNAMIC_EXECUTION]: verify_blocks.py performs network requests to the crates.io API to resolve dependency names and versions during the compilation check phase.

content-gen