The Agent Skills Directory

[PROMPT_INJECTION]: The skill contains standard instructional language focused on PHPStan rule development. No attempts to override agent behavior, bypass safety filters, or extract system prompts were identified.
[DATA_EXFILTRATION]: No sensitive file paths (such as SSH keys or AWS credentials) or network exfiltration patterns were found. The skill mentions local files like phpstan.neon and test fixtures, which are standard in the documented development workflow.
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or other secrets were detected within the documentation or code examples.
[REMOTE_CODE_EXECUTION]: There are no commands that download and execute remote scripts (e.g., curl | bash). Mentions of phpstan analyse and phpunit are provided as instructional context for testing code locally.
[OBFUSCATION]: No Base64-encoded strings, zero-width characters, homoglyphs, or other obfuscation techniques designed to hide malicious payloads were found.
[INDIRECT_PROMPT_INJECTION]: The skill acts as a documentation resource and does not ingest untrusted external data or provide exploitable tool capabilities that would enable indirect injection attacks.
[DYNAMIC_CONTEXT_INJECTION]: No use of the dynamic execution syntax (!command) was found in the SKILL.md file.

custom-rules