The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill analyzes TYPO3 extension source code (PHP), configurations (JSON, XML, YAML, TypoScript), and test suites, which creates an attack surface for instructions embedded in analyzed data.
Ingestion points: The agent reads composer.json, Classes/, Tests/, and Configuration/ directories as part of the audit and migration workflow.
Boundary markers: There are no explicit boundary markers or instructions to ignore embedded prompts when the agent reads external file content.
Capability inventory: The skill utilizes composer, php, git, and grep, and has the ability to write to the file system.
Sanitization: No sanitization is performed on the analyzed content before it is processed by the agent.
[DYNAMIC_EXECUTION]: The skill uses php -r to execute small local code snippets for parsing PHP tokens and analyzing composer.json metadata. This is used to verify that removed TYPO3 types are not used without imports.
Evidence: Found in SKILL.md step 9 and checkpoints.yaml (mechanisms TU-58 and TU-60).
[COMMAND_EXECUTION]: The skill relies on standard system and development tools to perform upgrades and verify extension state.
Evidence: Extensive use of composer, git, grep, sed, php, and phpunit across the instructions and troubleshooting guides.

typo3-extension-upgrade