The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes multiple shell commands for environment detection and project validation.
Evidence: Use of bunx tsc, node -v, and bun run typecheck in SKILL.md to interact with the local development environment.
Evidence: The scripts/ts_diagnostic.py script executes system commands including npx tsc, grep, and wc using subprocess.run.
[DYNAMIC_EXECUTION]: The skill utilizes dynamic code execution for diagnostic purposes.
Evidence: In SKILL.md, a node -e command is used to execute inline JavaScript that parses package.json to detect the tooling ecosystem.
Evidence: scripts/ts_diagnostic.py uses subprocess.run(shell=True), which allows for shell-interpreted command execution.
[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted project data, creating a surface for indirect prompt injection.
Ingestion points: The skill reads package.json, tsconfig.json, and project source files (via grep and direct file reads in SKILL.md and scripts/ts_diagnostic.py).
Boundary markers: Absent; there are no explicit delimiters or instructions to the model to treat the content of these files as untrusted.
Capability inventory: The skill can execute shell commands (tsc, vitest, grep, node, bun), read files, and analyze project structure.
Sanitization: Absent; the skill does not sanitize or validate the content of the project files before processing them or using them to inform subsequent agent actions.

typescript-expert