The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were detected. The skill defines a logical and safe framework for software debugging.
[DATA_EXPOSURE]: The skill includes example shell commands for diagnosing build and signing issues, such as checking for environment variable presence and listing keychain identities. These are standard developer troubleshooting steps. The instructions specifically demonstrate a safe pattern for checking secret availability (using shell parameter expansion) that avoids printing the actual secret values.
[INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process external error messages and stack traces, which are untrusted ingestion points (SKILL.md) that lack explicit boundary markers. The agent has capabilities for shell execution and file-writing to create tests and gather evidence. While no specific sanitization of external content is mentioned, the methodology's focus on systematic root-cause analysis and evidence verification acts as a strong behavioral safeguard against following potentially malicious instructions embedded in logs.

systematic-debugging