The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes project code, database schema analyzer outputs, and reports from automated tools (Rector, Fractor) as primary inputs. This creates a surface area where malicious instructions embedded in a compromised project's files or tool logs could influence the agent's actions during the upgrade process.
Ingestion points: Accesses and interprets project-specific files, registry/cache rebuild outputs, and schema analyzer data as specified in SKILL.md.
Boundary markers: While the instructions emphasize 'fixed-point proof' and 'identity/fingerprints', there are no explicit delimiters or instructions to ignore potential natural language commands embedded within the processed code or data.
Capability inventory: The skill utilizes file system writes, shell command execution via DDEV (TYPO3 CLI, Composer), and repository management via Git as described in the Invariants and Specialist Routes sections.
Sanitization: No explicit sanitization or validation of content from tool outputs or schema analysis is mentioned before being used to guide migration steps.
[COMMAND_EXECUTION]: The skill inherently requires the execution of shell commands to perform its core functions of software modernization and environment management.
Evidence: Routinely executes DDEV, Composer, Rector, Fractor, and TYPO3 commands for dependency resolution, API migration, and schema updates.
[EXTERNAL_DOWNLOADS]: The skill manages external dependencies through standard package management systems.
Evidence: Downloads and updates PHP packages including typo3/cms-core and the third-party package spooner/deployer-information via Composer.

typo3-upgrade-migration