The Agent Skills Directory

[DYNAMIC_CONTEXT_INJECTION]: The skill utilizes the dynamic context injection pattern (!command) in SKILL.md to execute a Node.js script when the skill is loaded. This script inspects the local project environment by reading package.json to extract Capacitor-related dependencies and verifying the presence of key native files like SceneDelegate.swift and Podfile. This is used to provide the agent with an immediate snapshot of the project's state.
[COMMAND_EXECUTION]: The migration process involves executing several shell commands for environment preflight checks (node, xcodebuild, java), automated migration via the Capacitor CLI (npx cap migrate), and native build processes (gradlew, xcodebuild, pod install). These operations are consistent with the skill's primary purpose of performing a major version upgrade.
[EXTERNAL_DOWNLOADS]: The instructions direct the agent to install and update numerous packages from the official NPM registry, specifically targeting the @next prerelease tags for Capacitor core and plugin libraries.
[INDIRECT_PROMPT_INJECTION]: The skill operates by ingesting and analyzing content from the local project repository, which constitutes an attack surface for indirect prompt injection if those files contain malicious instructions.
Ingestion points: The agent reads and processes local files including package.json, Info.plist, Podfile, build.gradle, and variables.gradle to identify required changes.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to distinguish between the untrusted content of the project files and the migration instructions.
Capability inventory: The skill possesses capabilities to execute shell commands, run Node.js scripts, install external packages, and perform native builds.
Sanitization: No sanitization or validation logic is present to filter the content of the project files before they are interpreted by the agent.

capacitor-app-upgrade-v8-to-v9