lc-cheatsheet
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external LeetCode problem data and instructions which are then used to generate and execute code snippets. * Ingestion points: User-supplied LeetCode numbers and existing repository content (SKILL.md Step 1). * Boundary markers: Not present in instructions. * Capability inventory: Execution of Bash, Python3, and Node.js commands (SKILL.md Steps 6, 7, 8). * Sanitization: No sanitization of ingested code or problem descriptions before execution.
- [DYNAMIC_EXECUTION]: Step 6 involves running generated Python snippets using `python3
- <<'PY'`. This creates an execution path for code that may be influenced by external or untrusted descriptions.
- [COMMAND_EXECUTION]: The skill utilizes several shell-based tools (grep, sed, node, bash, npm) to manipulate the filesystem and run build processes based on task parameters.
Audit Metadata