The Agent Skills Directory

[External Binary Execution]: The skill includes instructions to download a static binary from the OPA project's official GitHub repository to a temporary directory, make it executable, and run it to analyze IR plans. While this is a common workflow for compatibility analysis, executing binaries from external sources is a security consideration.
Evidence: SKILL.md and references/change-surface.md contain commands using curl to fetch the opa_linux_amd64_static binary, followed by chmod +x and execution for opa build and opa version commands.
[Processing of External Data]: The skill ingests and processes release notes and documentation from external web sources to generate compatibility reports. This represents a surface where the agent's reasoning and output are influenced by third-party content.
Ingestion points: GitHub releases pages and official OPA documentation sites (openpolicyagent.org).
Boundary markers: No explicit delimiters or instructions to ignore embedded commands in the external data are present.
Capability inventory: The skill utilizes file system operations (grep, rm, mkdir, tar), network requests (curl), and shell command execution (cmake, ctest, ninja, and the downloaded OPA binary).
Sanitization: The procedure does not specify explicit validation or sanitization steps for the fetched external content before it is used to generate report summaries and checklists.
[System Command Execution]: The skill performs several local shell operations, including project configuration with CMake, build execution with Ninja, and running test suites. These are standard development activities for the target repository.

opa-compat-check