The Agent Skills Directory

[SAFE]: No malicious patterns, obfuscation, or data exfiltration attempts were identified. The skill's instructions are consistent with its purpose of aiding PHP static analysis.
[COMMAND_EXECUTION]: The skill mentions the command ./vendor/bin/phpstan to execute the analysis. This targets a standard project-local binary and does not involve downloading or executing remote code.
[PROMPT_INJECTION]: The skill identifies an attack surface for indirect prompt injection as it instructs the agent to read project-level configuration files to tailor its behavior.
Ingestion points: The agent is instructed to read phpstan.neon, phpstan-baseline.neon, and custom rule files in src/Support/PHPStan/Rules/ (SKILL.md).
Boundary markers: The skill does not provide specific delimiters or warnings to ignore potentially malicious instructions embedded in the project files.
Capability inventory: The skill uses file-read access to project source and config files, and shell execution for the local PHPStan binary.
Sanitization: There is no mention of sanitizing or validating the contents of the configuration files before they are processed by the agent.

larastan