The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides instructions to create and run multiple shell scripts (package_app.sh, compile_and_run.sh, sign-and-notarize.sh) that execute system-level commands such as pkill, codesign, security, ditto, xcrun, git, and gh to manage the build, signing, and release cycle.
[DYNAMIC_EXECUTION]: In the 'Sparkle Updates' section, a script attempts to execute a binary located at ./bin/sign_update. The skill does not provide this binary or its source code, meaning it executes a pre-existing or externally provided executable at runtime.
[INDIRECT_PROMPT_INJECTION]: The skill's workflow depends on loading configuration from a version.env file.
Ingestion points: version.env (sourced in multiple scripts in SKILL.md).
Boundary markers: None.
Capability inventory: pkill, codesign, ditto, xcrun, gh.
Sanitization: None. Variables like $APP_NAME are used directly in shell commands (e.g., pkill -x "$APP_NAME"), which could lead to command injection if the agent processes a malicious environment file.
[PRIVILEGE_ESCALATION]: The skill suggests creating a development keychain using security create-keychain -p "" dev-signing.keychain. Using an empty password (-p "") for a keychain is a weak security practice that reduces the protection of cryptographic assets stored within it.

macos-spm-app-packaging