The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill is designed to read and refactor local TypeScript source code, which introduces a potential surface for indirect prompt injection if the files contain malicious comments or code designed to influence agent behavior.
Ingestion points: The skill ingests files matching *.ts, *.tsx, *.mts, and *.cts as defined in SKILL.md.
Boundary markers: There are no explicit delimiters or instructions to treat ingested code as data rather than instructions.
Capability inventory: The skill utilizes shell commands including npm run build, npm test, git log, find, and grep across multiple scripts.
Sanitization: No sanitization or content filtering is implemented for the files being refactored.
[COMMAND_EXECUTION]: The skill uses standard development tools and scripts to assess codebase state and verify refactoring steps.
Evidence: SKILL.md and references/smell-catalog.md contain instructions to execute npm run build, npm test, find, wc, git log, and grep. These commands are consistent with the skill's primary purpose of code refactoring and assessment.

do-typescript-refactoring