dig

Read-only debugging lets you investigate bugs and system behavior by gathering evidence from files and logs without applying changes. The agent stays in investigation mode, analyzing outputs and proposing hypotheses before any modifications occur.

Incident triage with a CLI agent involves guided troubleshooting that clarifies goals, gathers context from logs and configs, and proposes hypotheses. The agent maintains read-only operations by default to ensure safe investigation until root causes are identified.

Yes, root cause analysis can be performed in investigation-only mode where the CLI agent reads configs and logs to understand system behavior. It explicitly avoids write operations, requiring explicit single-use permission before modifying anything.

Permission-gated troubleshooting explains the exact change needed and asks for explicit, single-use approval before modifying files or configs. The agent maintains investigation mode by default and only proceeds with changes after receiving user permission.

A read-only investigation workflow is needed when you want to understand bugs, features, or system behavior before making any modifications. It ensures you can gather evidence and propose hypotheses safely, moving forward with confidence once root causes are clear.

Read-only debugging with CLI agents limits modifications until explicit permission is granted. The agent cannot apply fixes or write changes during investigation mode, meaning all troubleshooting stays analytical until the user approves specific single-use modifications.