The Agent Skills Directory

[PROMPT_INJECTION]: No evidence of instructions designed to bypass safety guidelines, override system behavior, or extract internal prompts was found. The skill maintains a consistent role as a technical reviewer.
[DATA_EXFILTRATION]: No network operations, credential usage, or access to sensitive file paths were detected. The skill operates locally on the provided code diffs.
[REMOTE_CODE_EXECUTION]: There are no patterns involving remote script downloads (e.g., curl|bash) or execution of untrusted external content.
[COMMAND_EXECUTION]: The skill does not attempt to execute shell commands, manage processes, or modify system configurations. The disable-model-invocation: true flag in the frontmatter further restricts tool usage.
[INDIRECT_PROMPT_INJECTION]: While the skill ingests external code for review (the attack surface), it lacks the capabilities (file writes, network requests, tool calls) to be exploited. The instructions focus on static analysis and generating a structured markdown table.
[OBFUSCATION]: No Base64, zero-width characters, or hidden payloads were found in the instructions or the reference file. External URLs for developer tools (easing.dev, easings.co) are provided in plain text for reference.

review-animations