The Agent Skills Directory

[SAFE]: The skill defines a rigorous debugging process that prioritizes evidence gathering and root cause identification over rapid, unverified fixes. This approach reduces the likelihood of introducing regressions or security vulnerabilities during the debugging process.
[SAFE]: Documentation in SKILL.md and defense-in-depth.md explicitly promotes security best practices, such as validating input at multiple architectural layers and avoiding the commitment of sensitive diagnostic data (secrets, PII) to version control.
[SAFE]: The utility script find-polluter.sh is a standard developer tool that uses npm test to isolate tests causing environmental side effects. It does not perform unauthorized network operations or privilege escalation.
[SAFE]: The TypeScript code in condition-based-waiting-example.ts provides benign utility functions for asynchronous polling in test environments, replacing unreliable arbitrary timeouts with condition-based logic.
[SAFE]: The included 'Pressure Test' scenarios are instructional content designed to evaluate and guide an AI's decision-making process during technical investigations. They do not contain executable malicious instructions or prompt injection attacks targeting the system's safety boundaries.

systematic-debugging