Examples

These examples show how coding-ethos should be used by humans and AI agents. They are intentionally small so the important path is visible: agents should ask the MCP server for policy and lint context instead of running raw tooling and guessing at remediation.

Available Examples

• MCP lint advice: run the local MCP server, inspect available tools, and use lint_check/lint_advice as the preferred agent path for static-analysis work.
• CEL policy: keep custom policy expressions attached to the ETHOS principle they enforce.
• GitHub SARIF CI: generated code-scanning gates and agent remediation flow.
• Agent command block: blocked unsafe command handling and MCP follow-up.
• Runtime sandbox: tool capabilities, sandbox evidence, and degraded-mode reporting.

Planned Examples

• Editor SARIF integration.
• Red-team regression authoring.
• Consumer repo cutover with generated agent settings.