The Agent Skills Directory

[SAFE]: The skill is entirely instructional, providing best practices and evaluation test cases for Golang development. Technical analysis of the content shows it to be high-quality, legitimate documentation for preventing software bugs.
[PROMPT_INJECTION]: No evidence of prompt injection or instructions to bypass safety filters was found. The use of 'IMPORTANT' or 'CRITICAL' in the text is strictly related to coding best practices.
[DATA_EXFILTRATION]: No commands for accessing sensitive local files (e.g., credentials, SSH keys) or environment variables are present. There are no network-bound operations aimed at external domains.
[REMOTE_CODE_EXECUTION]: The skill does not download or execute remote scripts. The installation instructions are empty, and the tool requirements are limited to standard development utilities (go, git, golangci-lint).
[INDIRECT_PROMPT_INJECTION]: The skill is designed to analyze user-provided Go code, which represents an ingestion surface for untrusted data. However, the skill provides specific defensive patterns for the agent to follow and contains no exploitable logic that would allow user data to be executed as instructions. Given its primary purpose as a code review tool, this risk is considered negligible. (Severity: LOW, downgraded to SAFE per primary purpose rule).

golang-safety