{ "name": "bug", "description": "Investigate a bug, prove it with a regression test, implement the fix, then verify adversarially.", "args": { "hint": "", "params": ["bug"] }, "schemas": { "Investigation": { "type": "object", "properties": { "summary": { "type": "string" }, "reproduction": { "type": "string" }, "likely_cause": { "type": "string" }, "files": { "type": "array", "items": { "type": "string" } } }, "required": ["summary", "reproduction", "likely_cause", "files"] }, "Regression": { "type": "object", "properties": { "ready": { "type": "boolean", "description": "true only when the regression test fails for the expected bug" }, "test_files": { "type": "array", "items": { "type": "string" } }, "command": { "type": "string" }, "failure": { "type": "string" } }, "required": ["ready", "test_files", "command", "failure"] }, "Fix": { "type": "object", "properties": { "fixed": { "type": "boolean", "description": "true only when the regression test and relevant checks pass" }, "summary": { "type": "string" }, "modified_files": { "type": "array", "items": { "type": "string" } }, "checks": { "type": "array", "items": { "type": "string" } } }, "required": ["fixed", "summary", "modified_files", "checks"] }, "Verdict": { "type": "object", "properties": { "dimension": { "type": "string" }, "passed": { "type": "boolean" }, "findings": { "type": "array", "items": { "type": "string" } } }, "required": ["dimension", "passed", "findings"] } }, "phases": [ { "id": "investigate", "kind": "single", "step": { "summary": "Investigate bug: {args.bug}", "prompt": "Investigate this bug without modifying files: {args.bug}. Reproduce or trace the failing path, inspect callers and shared routes, identify the likely root cause, and list relevant files. Use read/grep/find/bash as needed. Return JSON matching Investigation.", "tools": ["read", "grep", "find", "ls", "bash"], "model": "medium", "thinkingLevel": "high", "schema": "Investigation" } }, { "id": "regression", "kind": "single", "step": { "summary": "Create regression test: {args.bug}", "prompt": "Create the smallest regression test for this bug: {args.bug}. Investigation: {investigate.results}. Do not fix production code. Run the focused test and set ready=true only when it fails for the expected reason, not from a test error. Return JSON matching Regression.", "tools": ["read", "grep", "find", "ls", "bash", "edit", "write"], "model": "big", "thinkingLevel": "medium", "schema": "Regression" } }, { "id": "fix", "kind": "single", "when": "{regression.results | where ready}", "step": { "summary": "Fix root cause: {args.bug}", "prompt": "Fix the root cause of this bug: {args.bug}. Investigation: {investigate.results}. Regression evidence: {regression.results}. Prefer the shared route over caller-specific patches. Make the minimum production change, run the regression test and relevant checks, and set fixed=true only when they pass. Then return JSON matching Fix.", "tools": ["read", "grep", "find", "ls", "bash", "edit", "write"], "model": "big", "thinkingLevel": "medium", "schema": "Fix" } }, { "id": "verify", "kind": "fanout", "when": "{fix.results | where fixed}", "over": ["regression", "correctness", "security"], "step": { "summary": "Verify fix from {item} perspective", "prompt": "Adversarially verify the bug fix from the {item} perspective. Bug: {args.bug}. Investigation: {investigate.results}. Regression: {regression.results}. Fix: {fix.results}. Inspect the diff and run focused checks as needed; do not modify files. Set passed=false and list concrete findings if the fix is incomplete, the regression test is weak, or new risk was introduced. Echo dimension as {item}. Return JSON matching Verdict.", "tools": ["read", "grep", "find", "ls", "bash"], "model": "big", "thinkingLevel": "high", "schema": "Verdict" } } ], "return": "{verify.results}" }