{"meta":{"title":"Post-tool use hook","intro":"The onPostToolUse hook is called after a tool executes successfully. Use it to:","product":"GitHub Copilot","breadcrumbs":[{"href":"/en/copilot","title":"GitHub Copilot"},{"href":"/en/copilot/how-tos","title":"How-tos"},{"href":"/en/copilot/how-tos/copilot-sdk","title":"Copilot SDK"},{"href":"/en/copilot/how-tos/copilot-sdk/hooks","title":"Use hooks"},{"href":"/en/copilot/how-tos/copilot-sdk/hooks/post-tool-use","title":"Post Tool Use"}],"documentType":"article"},"body":"# Post-tool use hook\n\nThe onPostToolUse hook is called after a tool executes successfully. Use it to:\n\n\n\n\n\n* Transform or filter tool results\n* Log tool execution for auditing\n* Add context based on results\n* Suppress results from the conversation\n\n> **Failure variant** — `onPostToolUse` only fires for successful tool executions. To observe **failed** tool calls, register `onPostToolUseFailure` (`on_post_tool_use_failure` in Python, `OnPostToolUseFailure` in Go/.NET, `on_post_tool_use_failure` in Rust). The handler receives `{ sessionId, toolName, toolArgs, error, timestamp, workingDirectory }` — the `error` field is a string extracted from the tool's failure result — and may return `{ additionalContext: string }` to inject extra guidance for the model (e.g. retry hints). See the [Session hooks](/en/copilot/how-tos/copilot-sdk/hooks/hooks-overview) for the full list. \n\n## Hook signature\n\n
\n
TypeScript
\n\n```typescript\ntype PostToolUseHandler = (\n input: PostToolUseHookInput,\n invocation: HookInvocation,\n) => Promise;\n```\n\n
\n\n
Python
\n\n```python\nPostToolUseHandler = Callable[\n [PostToolUseHookInput, dict[str, str]],\n Awaitable[PostToolUseHookOutput | None]\n]\n```\n\n
\n\n
Go
\n\n```golang\ntype PostToolUseHandler func(\n input PostToolUseHookInput,\n invocation HookInvocation,\n) (*PostToolUseHookOutput, error)\n```\n\n
\n\n
.NET
\n\n```csharp\npublic delegate Task PostToolUseHandler(\n PostToolUseHookInput input,\n HookInvocation invocation);\n```\n\n
\n\n
Java
\n\n```java\n@FunctionalInterface\npublic interface PostToolUseHandler {\n CompletableFuture handle(\n PostToolUseHookInput input,\n HookInvocation invocation);\n}\n```\n\n
\n\n
\n\n## Input\n\n| Field | Type | Description |\n| ------------------ | ------------------ | -------------------------------------- |\n| `timestamp` | SDK timestamp type | When the hook was triggered |\n| `workingDirectory` | string | Current working directory |\n| `toolName` | string | Name of the tool that was called |\n| `toolArgs` | object | Arguments that were passed to the tool |\n| `toolResult` | object | Result returned by the tool |\n\n## Output\n\nReturn `null` or `undefined` to pass through the result unchanged. Otherwise, return an object with any of these fields:\n\n| Field | Type | Description |\n| ------------------- | ------- | -------------------------------------------- |\n| `modifiedResult` | object | Modified result to use instead of original |\n| `additionalContext` | string | Extra context injected into the conversation |\n| `suppressOutput` | boolean | If true, result won't appear in conversation |\n\n## Examples\n\n### Log all tool results\n\n
\n
TypeScript
\n\n```typescript\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input, invocation) => {\n console.log(`[${invocation.sessionId}] Tool: ${input.toolName}`);\n console.log(` Args: ${JSON.stringify(input.toolArgs)}`);\n console.log(` Result: ${JSON.stringify(input.toolResult)}`);\n return null; // Pass through unchanged\n },\n },\n});\n```\n\n
\n\n
Python
\n\n```python\nfrom copilot.session import PermissionHandler\n\nasync def on_post_tool_use(input_data, invocation):\n print(f\"[{invocation['session_id']}] Tool: {input_data['toolName']}\")\n print(f\" Args: {input_data['toolArgs']}\")\n print(f\" Result: {input_data['toolResult']}\")\n return None # Pass through unchanged\n\nsession = await client.create_session(on_permission_request=PermissionHandler.approve_all, hooks={\"on_post_tool_use\": on_post_tool_use})\n```\n\n
\n\n
Go
\n\n```golang\nsession, _ := client.CreateSession(context.Background(), &copilot.SessionConfig{\n Hooks: &copilot.SessionHooks{\n OnPostToolUse: func(input copilot.PostToolUseHookInput, inv copilot.HookInvocation) (*copilot.PostToolUseHookOutput, error) {\n fmt.Printf(\"[%s] Tool: %s\\n\", inv.SessionID, input.ToolName)\n fmt.Printf(\" Args: %v\\n\", input.ToolArgs)\n fmt.Printf(\" Result: %v\\n\", input.ToolResult)\n return nil, nil\n },\n },\n})\n```\n\n
\n\n
.NET
\n\n```csharp\nvar session = await client.CreateSessionAsync(new SessionConfig\n{\n Hooks = new SessionHooks\n {\n OnPostToolUse = (input, invocation) =>\n {\n Console.WriteLine($\"[{invocation.SessionId}] Tool: {input.ToolName}\");\n Console.WriteLine($\" Args: {input.ToolArgs}\");\n Console.WriteLine($\" Result: {input.ToolResult}\");\n return Task.FromResult(null);\n },\n },\n});\n```\n\n
\n\n
Java
\n\n\n\n```java\nimport com.github.copilot.*;\nimport com.github.copilot.rpc.*;\nimport java.util.concurrent.CompletableFuture;\n\nvar hooks = new SessionHooks()\n .setOnPostToolUse((input, invocation) -> {\n System.out.println(\"[\" + invocation.getSessionId() + \"] Tool: \" + input.getToolName());\n System.out.println(\" Args: \" + input.getToolArgs());\n System.out.println(\" Result: \" + input.getToolResult());\n return CompletableFuture.completedFuture(null);\n });\n\nvar session = client.createSession(\n new SessionConfig()\n .setOnPermissionRequest(PermissionHandler.APPROVE_ALL)\n .setHooks(hooks)\n).get();\n```\n\n
\n\n
\n\n### Redact sensitive data\n\n```typescript\nconst SENSITIVE_PATTERNS = [\n /api[_-]?key[\"\\s:=]+[\"']?[\\w-]+[\"']?/gi,\n /password[\"\\s:=]+[\"']?[\\w-]+[\"']?/gi,\n /secret[\"\\s:=]+[\"']?[\\w-]+[\"']?/gi,\n];\n\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input) => {\n if (typeof input.toolResult === \"string\") {\n let redacted = input.toolResult;\n for (const pattern of SENSITIVE_PATTERNS) {\n redacted = redacted.replace(pattern, \"[REDACTED]\");\n }\n\n if (redacted !== input.toolResult) {\n return { modifiedResult: redacted };\n }\n }\n return null;\n },\n },\n});\n```\n\n### Truncate large results\n\n```typescript\nconst MAX_RESULT_LENGTH = 10000;\n\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input) => {\n const resultStr = JSON.stringify(input.toolResult);\n\n if (resultStr.length > MAX_RESULT_LENGTH) {\n return {\n modifiedResult: {\n truncated: true,\n originalLength: resultStr.length,\n content: resultStr.substring(0, MAX_RESULT_LENGTH) + \"...\",\n },\n additionalContext: `Note: Result was truncated from ${resultStr.length} to ${MAX_RESULT_LENGTH} characters.`,\n };\n }\n return null;\n },\n },\n});\n```\n\n### Add context based on results\n\n```typescript\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input) => {\n // If a file read returned an error, add helpful context\n if (input.toolName === \"read_file\" && input.toolResult?.error) {\n return {\n additionalContext:\n \"Tip: If the file doesn't exist, consider creating it or checking the path.\",\n };\n }\n\n // If shell command failed, add debugging hint\n if (input.toolName === \"shell\" && input.toolResult?.exitCode !== 0) {\n return {\n additionalContext:\n \"The command failed. Check if required dependencies are installed.\",\n };\n }\n\n return null;\n },\n },\n});\n```\n\n### Filter error stack traces\n\n```typescript\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input) => {\n if (input.toolResult?.error && input.toolResult?.stack) {\n // Remove internal stack trace details\n return {\n modifiedResult: {\n error: input.toolResult.error,\n // Keep only first 3 lines of stack\n stack: input.toolResult.stack.split(\"\\n\").slice(0, 3).join(\"\\n\"),\n },\n };\n }\n return null;\n },\n },\n});\n```\n\n### Audit trail for compliance\n\n```typescript\ninterface AuditEntry {\n timestamp: Date;\n sessionId: string;\n toolName: string;\n args: unknown;\n result: unknown;\n success: boolean;\n}\n\nconst auditLog: AuditEntry[] = [];\n\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input, invocation) => {\n auditLog.push({\n timestamp: input.timestamp,\n sessionId: invocation.sessionId,\n toolName: input.toolName,\n args: input.toolArgs,\n result: input.toolResult,\n success: !input.toolResult?.error,\n });\n\n // Optionally persist to database/file\n await saveAuditLog(auditLog);\n\n return null;\n },\n },\n});\n```\n\n### Suppress noisy results\n\n```typescript\nconst NOISY_TOOLS = [\"list_directory\", \"search_codebase\"];\n\nconst session = await client.createSession({\n hooks: {\n onPostToolUse: async (input) => {\n if (NOISY_TOOLS.includes(input.toolName)) {\n // Summarize instead of showing full result\n const items = Array.isArray(input.toolResult)\n ? input.toolResult\n : input.toolResult?.items || [];\n\n return {\n modifiedResult: {\n summary: `Found ${items.length} items`,\n firstFew: items.slice(0, 5),\n },\n };\n }\n return null;\n },\n },\n});\n```\n\n## Best practices\n\n1. **Return `null` when no changes needed** - This is more efficient than returning an empty object or the same result.\n\n2. **Be careful with result modification** - Changing results can affect how the model interprets tool output. Only modify when necessary.\n\n3. **Use `additionalContext` for hints** - Instead of modifying results, add context to help the model interpret them.\n\n4. **Consider privacy when logging** - Tool results may contain sensitive data. Apply redaction before logging.\n\n5. **Keep hooks fast** - Post-tool hooks run synchronously. Heavy processing should be done asynchronously or batched.\n\n## See also\n\n* [Use hooks](/en/copilot/how-tos/copilot-sdk/hooks)\n* [Pre-tool use hook](/en/copilot/how-tos/copilot-sdk/hooks/pre-tool-use)\n* [Error handling hook](/en/copilot/how-tos/copilot-sdk/hooks/error-handling)"}