@code-yeongyu/senpi 2026.7.13 → 2026.7.14-3

package/docs/custom-provider.md

@@ -722,6 +722,8 @@ interface ProviderModelConfig {
     thinkingFormat?: "openai" | "openrouter" | "deepseek" | "together" | "zai" | "qwen" | "chat-template" | "qwen-chat-template" | "string-thinking" | "ant-ling";
     chatTemplateKwargs?: Record<string, string | number | boolean | null | { "$var": "thinking.enabled" | "thinking.effort"; omitWhenOff?: boolean }>;
     cacheControlFormat?: "anthropic";
+    sessionAffinityFormat?: "openai" | "openai-nosession" | "openrouter";
+    sendSessionAffinityHeaders?: boolean;
 
     // anthropic-messages
     supportsEagerToolInputStreaming?: boolean;

package/docs/models.md

@@ -447,6 +447,8 @@ For providers with partial OpenAI compatibility, use the `compat` field.
 | `thinkingFormat` | Use `reasoning_effort`, `openrouter`, `deepseek`, `together`, `zai`, `qwen`, `chat-template`, or `qwen-chat-template` thinking parameters |
 | `chatTemplateKwargs` | `chat_template_kwargs` values for `thinkingFormat: "chat-template"`; use `{ "$var": "thinking.enabled" }` or `{ "$var": "thinking.effort" }` for pi-controlled thinking values |
 | `cacheControlFormat` | Use Anthropic-style `cache_control` markers on the system prompt, last tool definition, and last user/assistant text content. Currently only `anthropic` is supported. |
+| `sendSessionAffinityHeaders` | For `openai-completions`, send session-affinity headers from the session id when caching is enabled. Default: `false`. |
+| `sessionAffinityFormat` | For `openai-completions` and `openai-responses`, the session-affinity header format: `openai` sends `session_id`/`x-client-request-id` (completions also `x-session-affinity`), `openai-nosession` omits the underscore-containing `session_id` header, `openrouter` sends `x-session-id`. Does not affect the `prompt_cache_key` body param. Default: auto-detected. |
 | `supportsStrictMode` | Include the `strict` field in tool definitions |
 | `supportsLongCacheRetention` | Whether the provider accepts long cache retention when cache retention is `long`: `prompt_cache_retention: "24h"` for OpenAI prompt caching, or `cache_control.ttl: "1h"` when `cacheControlFormat` is `anthropic`. Default: `true`. |
 | `openRouterRouting` | OpenRouter provider routing preferences. This object is sent as-is in the `provider` field of the [OpenRouter API request](https://openrouter.ai/docs/guides/routing/provider-selection). |

package/examples/AGENTS.md

@@ -0,0 +1,34 @@
+# packages/coding-agent/examples
+
+Runnable examples for the public Senpi SDK and extension API. Examples are documentation-quality code and must reflect shipped APIs, not private core internals. Unqualified paths below are relative to this directory; `packages/...` paths are repository-relative.
+
+## STRUCTURE
+
+```text
+extensions/        Tools, commands, UI, providers, hooks, resources
+extensions/*/      Multi-file examples and nested private workspaces
+sdk/               Programmatic SDK usage
+rpc-extension-ui.ts RPC-compatible extension UI example
+```
+
+## CONVENTIONS
+
+- Import public package surfaces such as `@code-yeongyu/senpi` and `@earendil-works/pi-ai`; do not reach into `packages/coding-agent/src/core/` internals.
+- Keep examples small enough to teach one pattern, while preserving real error, cleanup, cancellation, and persistence behavior where relevant.
+- Extension factories have no top-level runtime side effects. Register work through the public `pi.*` API and lifecycle events.
+- New interactive examples should use configurable keybindings and themed TUI helpers. Existing demos may keep fixed controls when the control scheme is part of the example. Direct terminal writes belong only in examples explicitly teaching a terminal protocol; ordinary SDK examples may use normal stdout.
+- Tool string enums use the shared `StringEnum` helper for provider compatibility.
+- Stateful examples persist reconstructable state in session entries or tool-result details so fork/resume behavior remains valid.
+- Nested example packages are private workspaces with exact-pinned dependencies. Treat their manifests and lock impact as production dependency changes.
+
+## DOCUMENTATION CONTRACT
+
+- Keep `extensions/README.md`, `packages/coding-agent/docs/extensions.md`, and `packages/coding-agent/docs/sdk.md` aligned with public API changes.
+- New public extension capabilities should include a focused example when usage is not obvious from types alone.
+- Do not present experimental or internal behavior as stable API.
+
+## VALIDATION
+
+- Run the focused tests for the public API demonstrated by the example.
+- Typecheck examples through root `npm run check`.
+- Interactive examples require real CLI or visual QA when their behavior changes.

package/examples/extensions/custom-provider-anthropic/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "pi-extension-custom-provider",
-  "version": "0.80.6",
+  "version": "0.80.7",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "pi-extension-custom-provider",
-      "version": "0.80.6",
+      "version": "0.80.7",
       "dependencies": {
         "@anthropic-ai/sdk": "^0.52.0"
       }

package/examples/extensions/custom-provider-anthropic/package.json

@@ -1,7 +1,7 @@
 {
   "name": "pi-extension-custom-provider-anthropic",
   "private": true,
-  "version": "0.80.6",
+  "version": "0.80.7",
   "type": "module",
   "scripts": {
     "clean": "echo 'nothing to clean'",

package/examples/extensions/custom-provider-gitlab-duo/package.json

@@ -1,7 +1,7 @@
 {
   "name": "pi-extension-custom-provider-gitlab-duo",
   "private": true,
-  "version": "0.80.6",
+  "version": "0.80.7",
   "type": "module",
   "scripts": {
     "clean": "echo 'nothing to clean'",

package/examples/extensions/gondolin/package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "pi-extension-gondolin",
-	"version": "0.80.6",
+	"version": "0.80.7",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "pi-extension-gondolin",
-			"version": "0.80.6",
+			"version": "0.80.7",
 			"dependencies": {
 				"@earendil-works/gondolin": "0.12.0"
 			}

package/examples/extensions/gondolin/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "pi-extension-gondolin",
 	"private": true,
-	"version": "0.80.6",
+	"version": "0.80.7",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/sandbox/package-lock.json

@@ -1,12 +1,12 @@
 {
 	"name": "pi-extension-sandbox",
-	"version": "1.10.6",
+	"version": "1.10.7",
 	"lockfileVersion": 3,
 	"requires": true,
 	"packages": {
 		"": {
 			"name": "pi-extension-sandbox",
-			"version": "1.10.6",
+			"version": "1.10.7",
 			"dependencies": {
 				"@anthropic-ai/sandbox-runtime": "^0.0.26"
 			}

package/examples/extensions/sandbox/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "pi-extension-sandbox",
 	"private": true,
-	"version": "1.10.6",
+	"version": "1.10.7",
 	"type": "module",
 	"scripts": {
 		"clean": "echo 'nothing to clean'",

package/examples/extensions/with-deps/package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "pi-extension-with-deps",
-  "version": "0.80.6",
+  "version": "0.80.7",
   "lockfileVersion": 3,
   "requires": true,
   "packages": {
     "": {
       "name": "pi-extension-with-deps",
-      "version": "0.80.6",
+      "version": "0.80.7",
       "dependencies": {
         "ms": "^2.1.3"
       },

package/examples/extensions/with-deps/package.json

@@ -1,7 +1,7 @@
 {
   "name": "pi-extension-with-deps",
   "private": true,
-  "version": "0.80.6",
+  "version": "0.80.7",
   "type": "module",
   "scripts": {
     "clean": "echo 'nothing to clean'",

package/node_modules/@code-yeongyu/senpi-codemode/CHANGELOG.md

@@ -1,5 +1,31 @@
 # Changelog
 
+## [2026.7.14-3] - 2026-07-14
+
+### Added
+
+### Changed
+
+### Fixed
+
+## [2026.7.14-2] - 2026-07-14
+
+### Added
+
+### Changed
+
+### Fixed
+
+## [2026.7.14] - 2026-07-14
+
+### Added
+
+### Changed
+
+- Improved the `eval` prompt instructions and reuse-chain examples to teach persistent-state reuse, batch file processing, and parallel session-tool fan-out within a single cell.
+
+### Fixed
+
 ## [2026.7.13] - 2026-07-13
 
 ### Added

package/node_modules/@code-yeongyu/senpi-codemode/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@code-yeongyu/senpi-codemode",
-	"version": "2026.7.13",
+	"version": "2026.7.14-3",
 	"description": "Source-only senpi extension package for codemode evaluation tools",
 	"type": "module",
 	"main": "./src/index.ts",
@@ -30,14 +30,14 @@
 	},
 	"dependencies": {
 		"@babel/parser": "7.29.7",
-		"@earendil-works/pi-ai": "^2026.7.13",
+		"@earendil-works/pi-ai": "^2026.7.14-3",
 		"typebox": "1.1.38"
 	},
 	"peerDependencies": {
 		"@code-yeongyu/senpi": "*"
 	},
 	"devDependencies": {
-		"@code-yeongyu/senpi": "2026.7.13"
+		"@code-yeongyu/senpi": "2026.7.14-3"
 	},
 	"keywords": [
 		"senpi",

package/node_modules/@code-yeongyu/senpi-codemode/src/prompt/eval-prompt.ts

@@ -25,37 +25,28 @@ type EvalPromptExample = {
 	readonly code: string;
 };
 
-// senpi ToolDefinition has no examples field, so description embeds parity examples.
+// senpi ToolDefinition has no examples field, so description embeds the examples.
+// ADAPTATION: payloads diverge from omp's json-config chain to teach batch read,
+// comprehension filtering, and parallel tool.<name> fan-out while keeping the
+// three-cell reuse narrative.
 const REUSE_CHAIN_EXAMPLES = [
 	{
 		caption: "First call — set up once",
 		language: "py",
-		title: "imports",
-		code: "import json\nfrom pathlib import Path",
+		title: "collect targets",
+		code: "from pathlib import Path\nfrom collections import Counter\nfiles = [p for p in Path('src').rglob('*.ts') if 'test' not in p.parts]\nprint(len(files))",
 	},
 	{
-		caption: "Second call — reuse, do NOT re-import",
+		caption: "Second call — reuse `files`, batch-read in one cell",
 		language: "py",
-		title: "load config",
-		code: "data = json.loads(read('package.json'))\ndisplay(data)",
+		title: "scan usages",
+		code: "hits = Counter()\nfor p in files:\n    hits[p.name] = read(p).count('legacyClient')\ndisplay({k: v for k, v in hits.items() if v})",
 	},
 	{
-		caption: "Third call — reuse the loaded config",
+		caption: "Third call — reuse results, fan out session tools in parallel",
 		language: "py",
-		title: "scan deps",
-		code: "display(sorted(data['dependencies']))",
-	},
-	{
-		caption: "Ruby first call — set up once",
-		language: "rb",
-		title: "setup",
-		code: "require 'json'\npkg_path = 'package.json'",
-	},
-	{
-		caption: "Ruby second call — reuse, do NOT re-require",
-		language: "rb",
-		title: "load config",
-		code: "pkg = JSON.parse(read(pkg_path))\ndisplay(pkg.keys.sort)",
+		title: "confirm callsites",
+		code: "dirs = ['src/core', 'src/tools']\ndisplay(parallel([lambda d=d: tool.grep({'pattern': 'legacyClient', 'path': d}) for d in dirs]))",
 	},
 ] as const satisfies readonly EvalPromptExample[];
 
@@ -64,7 +55,9 @@ const EVAL_PROMPT_TEMPLATE = `Run one step of code in a persistent kernel.
 <instruction>
 **One eval call = one cell = one logical step.** State persists per language across separate eval calls and tool calls{{#if spawns}}, and \`task\` subagents{{/if}} — define helpers, datasets, and clients in one call, then later calls reuse them directly.
 
-Work incrementally: imports in one call, define in the next, test, then use — each its own eval call. Re-run setup ONLY after \`reset\`, a kernel crash, or a \`NameError\`/\`ReferenceError\` proving the state is gone. Parallelize work *within* a cell with the \`parallel(thunks)\` helper, not by batching steps.
+Work incrementally: imports in one call, define in the next, test, then use — each its own eval call. Re-run setup ONLY after \`reset\`, a kernel crash, or a \`NameError\`/\`ReferenceError\` proving the state is gone.
+
+**Batch through eval.** One cell replaces many one-shot tool calls: loop or comprehend over file sets with \`read()\`/stdlib instead of reading files one call at a time, post-process \`tool.<name>()\` results programmatically, and fan independent calls through \`parallel(thunks)\` within the cell — never one call per turn.
 
 Fields:
 
@@ -125,7 +118,7 @@ Pipe handles through stage helpers to build a dependency graph — acyclic waves
 {{/if}}
 
 <critical>
-Prior top-level names (\`data\`, \`sessions\`, helpers, imports) survive into the next eval call — reuse them; NEVER re-import, re-require, or re-declare a helper. Re-read a file only if it may have changed since the last read. Re-run setup only after \`reset\`, a crash, or a \`NameError\`/\`ReferenceError\`.
+Prior top-level names (\`data\`, \`sessions\`, helpers, imports) survive into the next eval call — reuse them; NEVER re-import, re-require, or re-declare a helper. Re-read a file only if it may have changed since the last read.
 </critical>`;
 
 export function buildEvalPrompt(

package/node_modules/@earendil-works/pi-agent-core/package.json

@@ -1,7 +1,7 @@
 {
 	"name": "@earendil-works/pi-agent-core",
 	"private": true,
-	"version": "2026.7.13",
+	"version": "2026.7.14-3",
 	"description": "General-purpose agent with transport abstraction, state management, and attachment support",
 	"type": "module",
 	"main": "./dist/index.js",
@@ -30,7 +30,7 @@
 		"prepublishOnly": "npm run clean && npm run build"
 	},
 	"dependencies": {
-		"@earendil-works/pi-ai": "^2026.7.13",
+		"@earendil-works/pi-ai": "^2026.7.14-3",
 		"ignore": "7.0.5",
 		"typebox": "1.1.38",
 		"yaml": "2.9.0"

package/node_modules/@earendil-works/pi-ai/README.md

@@ -1070,7 +1070,8 @@ interface OpenAICompletionsCompat {
   supportsReasoningEffort?: boolean; // Whether provider supports `reasoning_effort` (default: true)
   supportsUsageInStreaming?: boolean; // Whether provider supports `stream_options: { include_usage: true }` (default: true)
   supportsStrictMode?: boolean;      // Whether provider supports `strict` in tool definitions (default: true)
-  sendSessionAffinityHeaders?: boolean; // Whether to send `session_id`, `x-client-request-id`, and `x-session-affinity` from `sessionId` when caching is enabled (default: false)
+  sendSessionAffinityHeaders?: boolean; // Send session-affinity data from `sessionId` (default: false)
+  sessionAffinityFormat?: 'openai' | 'openai-nosession' | 'openrouter'; // Format for session affinity: 'openai' uses `prompt_cache_key`, `session_id`, `x-client-request-id`, and `x-session-affinity`; 'openai-nosession' uses `prompt_cache_key`, `x-client-request-id`, and `x-session-affinity`; 'openrouter' uses `x-session-id` (default: auto-detected)
   maxTokensField?: 'max_completion_tokens' | 'max_tokens';  // Which field name to use (default: max_completion_tokens)
   requiresToolResultName?: boolean;  // Whether tool results require the `name` field (default: false)
   requiresAssistantAfterToolResult?: boolean; // Whether tool results must be followed by an assistant message (default: false)
@@ -1085,7 +1086,7 @@ interface OpenAICompletionsCompat {
 
 interface OpenAIResponsesCompat {
   supportsDeveloperRole?: boolean;   // Whether provider supports `developer` role vs `system` (default: true)
-  sendSessionIdHeader?: boolean;     // Whether to send `session_id` from `sessionId` when caching is enabled (default: true)
+  sessionAffinityFormat?: 'openai' | 'openai-nosession' | 'openrouter'; // Session-affinity header format: 'openai' sends `session_id` and `x-client-request-id`; 'openai-nosession' sends `x-client-request-id`; 'openrouter' sends `x-session-id`. Does not affect the `prompt_cache_key` body param (default: auto-detected)
   supportsLongCacheRetention?: boolean; // Whether provider supports `prompt_cache_retention: "24h"` (default: true)
 }
 ```

package/node_modules/@earendil-works/pi-ai/dist/api/anthropic-messages.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"anthropic-messages.d.ts","sourceRoot":"","sources":["../../src/api/anthropic-messages.ts"],"names":[],"mappings":"AAAA,OAAO,SAAS,MAAM,mBAAmB,CAAC;AAU1C,OAAO,KAAK,EAWX,mBAAmB,EAEnB,cAAc,EACd,aAAa,EAMb,MAAM,aAAa,CAAC;AAgKrB,MAAM,MAAM,eAAe,GAAG,KAAK,GAAG,QAAQ,GAAG,MAAM,GAAG,OAAO,GAAG,KAAK,CAAC;AAE1E,MAAM,MAAM,wBAAwB,GAAG,YAAY,GAAG,SAAS,CAAC;AAuDhE,MAAM,WAAW,gBAAiB,SAAQ,aAAa;IACtD;;;;;;OAMG;IACH,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B;;;;OAIG;IACH,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B;;;;;;;;;;;OAWG;IACH,MAAM,CAAC,EAAE,eAAe,CAAC;IACzB;;;;;;;;;;;OAWG;IACH,eAAe,CAAC,EAAE,wBAAwB,CAAC;IAC3C;;;;;OAKG;IACH,mBAAmB,CAAC,EAAE,OAAO,CAAC;IAC9B;;;;OAIG;IACH,UAAU,CAAC,EAAE,MAAM,GAAG,KAAK,GAAG,MAAM,GAAG;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,CAAC;IACtE;;;;OAIG;IACH,MAAM,CAAC,EAAE,SAAS,CAAC;CACnB;AAkiBD,eAAO,MAAM,MAAM,EAAE,cAAc,CAAC,oBAAoB,EAAE,gBAAgB,CAoTzE,CAAC;AAoEF,eAAO,MAAM,YAAY,EAAE,cAAc,CAAC,oBAAoB,EAAE,mBAAmB,CA2ClF,CAAC"}
+{"version":3,"file":"anthropic-messages.d.ts","sourceRoot":"","sources":["../../src/api/anthropic-messages.ts"],"names":[],"mappings":"AAAA,OAAO,SAAS,MAAM,mBAAmB,CAAC;AAU1C,OAAO,KAAK,EAWX,mBAAmB,EAEnB,cAAc,EACd,aAAa,EAMb,MAAM,aAAa,CAAC;AAgKrB,MAAM,MAAM,eAAe,GAAG,KAAK,GAAG,QAAQ,GAAG,MAAM,GAAG,OAAO,GAAG,KAAK,CAAC;AAE1E,MAAM,MAAM,wBAAwB,GAAG,YAAY,GAAG,SAAS,CAAC;AAuDhE,MAAM,WAAW,gBAAiB,SAAQ,aAAa;IACtD;;;;;;OAMG;IACH,eAAe,CAAC,EAAE,OAAO,CAAC;IAC1B;;;;OAIG;IACH,oBAAoB,CAAC,EAAE,MAAM,CAAC;IAC9B;;;;;;;;;;;OAWG;IACH,MAAM,CAAC,EAAE,eAAe,CAAC;IACzB;;;;;;;;;;;OAWG;IACH,eAAe,CAAC,EAAE,wBAAwB,CAAC;IAC3C;;;;;OAKG;IACH,mBAAmB,CAAC,EAAE,OAAO,CAAC;IAC9B;;;;OAIG;IACH,UAAU,CAAC,EAAE,MAAM,GAAG,KAAK,GAAG,MAAM,GAAG;QAAE,IAAI,EAAE,MAAM,CAAC;QAAC,IAAI,EAAE,MAAM,CAAA;KAAE,CAAC;IACtE;;;;OAIG;IACH,MAAM,CAAC,EAAE,SAAS,CAAC;CACnB;AAghBD,eAAO,MAAM,MAAM,EAAE,cAAc,CAAC,oBAAoB,EAAE,gBAAgB,CAsTzE,CAAC;AAoEF,eAAO,MAAM,YAAY,EAAE,cAAc,CAAC,oBAAoB,EAAE,mBAAmB,CA2ClF,CAAC"}

package/node_modules/@earendil-works/pi-ai/dist/api/anthropic-messages.js

@@ -200,24 +200,6 @@ const REPLAYABLE_ANTHROPIC_PROVIDER_NATIVE_TYPES = new Set([
 function isReplayableAnthropicProviderNativeBlock(raw) {
     return isRecord(raw) && typeof raw.type === "string" && REPLAYABLE_ANTHROPIC_PROVIDER_NATIVE_TYPES.has(raw.type);
 }
-function stripEncryptedContentFromWebSearchResult(raw) {
-    const { encrypted_content: _encryptedContent, ...rest } = raw;
-    return rest;
-}
-function sanitizeReplayableAnthropicProviderNativeBlock(raw) {
-    if (!isReplayableAnthropicProviderNativeBlock(raw))
-        return undefined;
-    if (!isRecord(raw) || raw.type !== "web_search_tool_result" || !Array.isArray(raw.content))
-        return raw;
-    return {
-        ...raw,
-        content: raw.content.map((item) => {
-            if (!isRecord(item) || item.type !== "web_search_result")
-                return item;
-            return stripEncryptedContentFromWebSearchResult(item);
-        }),
-    };
-}
 function isSameAnthropicModel(message, model) {
     return message.provider === model.provider && message.api === model.api && message.model === model.id;
 }
@@ -837,25 +819,27 @@ export const stream = (model, context, options) => {
                     }
                     // Only update usage fields if present (not null).
                     // Preserves input_tokens from message_start when proxies omit it in message_delta.
-                    if (event.usage.input_tokens != null) {
-                        output.usage.input = event.usage.input_tokens;
-                    }
-                    if (event.usage.output_tokens != null) {
-                        output.usage.output = event.usage.output_tokens;
-                    }
-                    if (event.usage.cache_read_input_tokens != null) {
-                        output.usage.cacheRead = event.usage.cache_read_input_tokens;
-                    }
-                    if (event.usage.cache_creation_input_tokens != null) {
-                        output.usage.cacheWrite = event.usage.cache_creation_input_tokens;
-                    }
-                    // Anthropic reports reasoning tokens in `output_tokens_details.thinking_tokens` on the
-                    // final message_delta usage (a subset of output_tokens). SDK 0.91.1 omits the field from
-                    // its Usage type, so read it through a narrow cast. Verified against the live API.
-                    const thinkingTokens = event.usage
-                        .output_tokens_details?.thinking_tokens;
-                    if (thinkingTokens != null) {
-                        output.usage.reasoning = thinkingTokens;
+                    if (event.usage) {
+                        if (event.usage.input_tokens != null) {
+                            output.usage.input = event.usage.input_tokens;
+                        }
+                        if (event.usage.output_tokens != null) {
+                            output.usage.output = event.usage.output_tokens;
+                        }
+                        if (event.usage.cache_read_input_tokens != null) {
+                            output.usage.cacheRead = event.usage.cache_read_input_tokens;
+                        }
+                        if (event.usage.cache_creation_input_tokens != null) {
+                            output.usage.cacheWrite = event.usage.cache_creation_input_tokens;
+                        }
+                        // Anthropic reports reasoning tokens in `output_tokens_details.thinking_tokens` on the
+                        // final message_delta usage (a subset of output_tokens). SDK 0.91.1 omits the field from
+                        // its Usage type, so read it through a narrow cast. Verified against the live API.
+                        const thinkingTokens = event.usage
+                            .output_tokens_details?.thinking_tokens;
+                        if (thinkingTokens != null) {
+                            output.usage.reasoning = thinkingTokens;
+                        }
                     }
                     // Anthropic doesn't provide total_tokens, compute from components
                     output.usage.totalTokens =
@@ -1336,11 +1320,8 @@ function convertMessages(transformedMessages, model, isOAuthToken, cacheControl,
                     });
                 }
                 else if (block.type === "providerNative") {
-                    if (isSameModel) {
-                        const replayableBlock = sanitizeReplayableAnthropicProviderNativeBlock(block.raw);
-                        if (replayableBlock)
-                            blocks.push(replayableBlock);
-                    }
+                    if (isSameModel && isReplayableAnthropicProviderNativeBlock(block.raw))
+                        blocks.push(block.raw);
                 }
             }
             if (blocks.length === 0)