The Agent Skills Directory

[Dynamic Execution] (MEDIUM): The skill utilizes bun build --compile to generate binary executables from source code. Specifically, the 'Post-build' section instructs the agent to execute the resulting binary (e.g., ./app --help). This represents runtime compilation and execution of dynamically generated artifacts.
[Indirect Prompt Injection] (LOW): This skill presents an attack surface where malicious instructions could be embedded in the source data.
Ingestion points: The <entry> file (e.g., index.ts) is read and processed by the Bun compiler.
Boundary markers: No specific boundary markers or instructions to ignore embedded prompts are provided during the build process.
Capability inventory: The skill has access to the Bash tool, enabling it to execute the build command and the resulting binary.
Sanitization: There is no evidence of sanitization or validation of the input source file before it is compiled and executed.
[Command Execution] (LOW): The skill is designed to run arbitrary shell commands via the Bash tool to perform builds and verify outputs. This is a powerful capability that, while intended, relies on the integrity of the environment and the inputs provided to the agent.

bun-build