The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses powershell.exe to execute a build script (epf-build.ps1). This script is located outside the skill's immediate directory using a relative path (${CLAUDE_SKILL_DIR}/../epf-build/scripts/).
[CREDENTIALS_UNSAFE]: The skill's instructions and examples include a -Password parameter for database authentication. Passing sensitive credentials directly through command-line arguments is unsafe because they can be captured by system monitoring tools, process explorers, or shell history logs.
[DYNAMIC_EXECUTION]: The skill dynamically constructs a PowerShell command line by interpolating multiple user-provided arguments (e.g., -SourceFile, -OutputFile, -V8Path, -AdditionalV8Arguments) into a shell execution string.
[INDIRECT_PROMPT_INJECTION]: The skill reads and processes external files, including .v8-project.json for configuration and various XML files as source code. There is a risk that maliciously crafted content within these files could influence the agent's behavior during the build process.
Ingestion points: Reads .v8-project.json and XML source files defined by -SourceFile.
Boundary markers: None identified in the skill instructions to delimit untrusted file content.
Capability inventory: The skill has the capability to execute shell commands (Bash, powershell.exe) and read/glob files.
Sanitization: The instructions do not specify sanitization or validation routines for the data read from configuration or source files before processing.

erf-build