From d0a4cce41e6d61f0670da461ddd71cef65bc3d0a Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:05:08 -0500
Subject: [PATCH 1/6] feat: add dynamic template variables to
 messages.responsePrefix (#923)
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Adds support for template variables in `messages.responsePrefix` that
resolve dynamically at runtime with the actual model used (including
after fallback).

Supported variables (case-insensitive):
- {model} - short model name (e.g., "claude-opus-4-5", "gpt-4o")
- {modelFull} - full model identifier (e.g., "anthropic/claude-opus-4-5")
- {provider} - provider name (e.g., "anthropic", "openai")
- {thinkingLevel} or {think} - thinking level ("high", "low", "off")
- {identity.name} or {identityName} - agent identity name

Example: "[{model} | think:{thinkingLevel}]" → "[claude-opus-4-5 | think:high]"

Variables show the actual model used after fallback, not the intended
model. Unresolved variables remain as literal text.

Implementation:
- New module: src/auto-reply/reply/response-prefix-template.ts
- Template interpolation in normalize-reply.ts via context provider
- onModelSelected callback in agent-runner-execution.ts
- Updated all 6 provider message handlers (web, signal, discord,
  telegram, slack, imessage)
- 27 unit tests covering all variables and edge cases
- Documentation in docs/gateway/configuration.md and JSDoc

Fixes #923
---
 docs/gateway/configuration.md                 |  25 +++
 src/agents/identity.ts                        |   5 +
 .../reply/agent-runner-execution.ts           |   8 +
 src/auto-reply/reply/normalize-reply.ts       |  17 +-
 src/auto-reply/reply/reply-dispatcher.ts      |  15 +-
 .../reply/response-prefix-template.test.ts    | 181 ++++++++++++++++++
 .../reply/response-prefix-template.ts         |  97 ++++++++++
 src/auto-reply/types.ts                       |  10 +
 src/config/types.messages.ts                  |  15 +-
 .../monitor/message-handler.process.ts        |  21 ++
 src/imessage/monitor/monitor-provider.ts      |  26 ++-
 src/signal/monitor/event-handler.ts           |  26 ++-
 src/slack/monitor/message-handler/dispatch.ts |  21 ++
 src/telegram/bot-message-dispatch.ts          |  21 +-
 src/web/auto-reply/monitor/process-message.ts |  21 +-
 15 files changed, 500 insertions(+), 9 deletions(-)
 create mode 100644 src/auto-reply/reply/response-prefix-template.test.ts
 create mode 100644 src/auto-reply/reply/response-prefix-template.ts

diff --git a/docs/gateway/configuration.md b/docs/gateway/configuration.md
index 012bec824..9906e94b1 100644
--- a/docs/gateway/configuration.md
+++ b/docs/gateway/configuration.md
@@ -1184,6 +1184,31 @@ streaming, final replies) across channels unless already present.
 If `messages.responsePrefix` is unset, no prefix is applied by default.
 Set it to `"auto"` to derive `[{identity.name}]` for the routed agent (when set).
 
+#### Template variables
+
+The `responsePrefix` string can include template variables that resolve dynamically:
+
+| Variable | Description | Example |
+|----------|-------------|---------|
+| `{model}` | Short model name | `claude-opus-4-5`, `gpt-4o` |
+| `{modelFull}` | Full model identifier | `anthropic/claude-opus-4-5` |
+| `{provider}` | Provider name | `anthropic`, `openai` |
+| `{thinkingLevel}` | Current thinking level | `high`, `low`, `off` |
+| `{identity.name}` | Agent identity name | (same as `"auto"` mode) |
+
+Variables are case-insensitive (`{MODEL}` = `{model}`). `{think}` is an alias for `{thinkingLevel}`.
+Unresolved variables remain as literal text.
+
+```json5
+{
+  messages: {
+    responsePrefix: "[{model} | think:{thinkingLevel}]"
+  }
+}
+```
+
+Example output: `[claude-opus-4-5 | think:high] Here's my response...`
+
 WhatsApp inbound prefix is configured via `channels.whatsapp.messagePrefix` (deprecated:
 `messages.messagePrefix`). Default stays **unchanged**: `"[clawdbot]"` when
 `channels.whatsapp.allowFrom` is empty, otherwise `""` (no prefix). When using
diff --git a/src/agents/identity.ts b/src/agents/identity.ts
index 6ac80da2f..b78e2746d 100644
--- a/src/agents/identity.ts
+++ b/src/agents/identity.ts
@@ -26,6 +26,11 @@ export function resolveIdentityNamePrefix(
   return `[${name}]`;
 }
 
+/** Returns just the identity name (without brackets) for template context. */
+export function resolveIdentityName(cfg: ClawdbotConfig, agentId: string): string | undefined {
+  return resolveAgentIdentity(cfg, agentId)?.name?.trim() || undefined;
+}
+
 export function resolveMessagePrefix(
   cfg: ClawdbotConfig,
   agentId: string,
diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index fbc7a7451..d1f0e276d 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -125,6 +125,14 @@ export async function runAgentTurnWithFallback(params: {
           resolveAgentIdFromSessionKey(params.followupRun.run.sessionKey),
         ),
         run: (provider, model) => {
+          // Notify that model selection is complete (including after fallback).
+          // This allows responsePrefix template interpolation with the actual model.
+          params.opts?.onModelSelected?.({
+            provider,
+            model,
+            thinkLevel: params.followupRun.run.thinkLevel,
+          });
+
           if (isCliProvider(provider, params.followupRun.run.config)) {
             const startedAt = Date.now();
             emitAgentEvent({
diff --git a/src/auto-reply/reply/normalize-reply.ts b/src/auto-reply/reply/normalize-reply.ts
index f46a49b9f..ac8e85f2e 100644
--- a/src/auto-reply/reply/normalize-reply.ts
+++ b/src/auto-reply/reply/normalize-reply.ts
@@ -1,9 +1,15 @@
 import { stripHeartbeatToken } from "../heartbeat.js";
 import { HEARTBEAT_TOKEN, isSilentReplyText, SILENT_REPLY_TOKEN } from "../tokens.js";
 import type { ReplyPayload } from "../types.js";
+import {
+  resolveResponsePrefixTemplate,
+  type ResponsePrefixContext,
+} from "./response-prefix-template.js";
 
 export type NormalizeReplyOptions = {
   responsePrefix?: string;
+  /** Context for template variable interpolation in responsePrefix */
+  responsePrefixContext?: ResponsePrefixContext;
   onHeartbeatStrip?: () => void;
   stripHeartbeat?: boolean;
   silentToken?: string;
@@ -36,13 +42,18 @@ export function normalizeReplyPayload(
     text = stripped.text;
   }
 
+  // Resolve template variables in responsePrefix if context is provided
+  const effectivePrefix = opts.responsePrefixContext
+    ? resolveResponsePrefixTemplate(opts.responsePrefix, opts.responsePrefixContext)
+    : opts.responsePrefix;
+
   if (
-    opts.responsePrefix &&
+    effectivePrefix &&
     text &&
     text.trim() !== HEARTBEAT_TOKEN &&
-    !text.startsWith(opts.responsePrefix)
+    !text.startsWith(effectivePrefix)
   ) {
-    text = `${opts.responsePrefix} ${text}`;
+    text = `${effectivePrefix} ${text}`;
   }
 
   return { ...payload, text };
diff --git a/src/auto-reply/reply/reply-dispatcher.ts b/src/auto-reply/reply/reply-dispatcher.ts
index 5cc3a97fa..b8eeca477 100644
--- a/src/auto-reply/reply/reply-dispatcher.ts
+++ b/src/auto-reply/reply/reply-dispatcher.ts
@@ -1,6 +1,7 @@
 import type { HumanDelayConfig } from "../../config/types.js";
 import type { GetReplyOptions, ReplyPayload } from "../types.js";
 import { normalizeReplyPayload } from "./normalize-reply.js";
+import type { ResponsePrefixContext } from "./response-prefix-template.js";
 import type { TypingController } from "./typing.js";
 
 export type ReplyDispatchKind = "tool" | "block" | "final";
@@ -33,6 +34,11 @@ const sleep = (ms: number) => new Promise((resolve) => setTimeout(resolve, ms));
 export type ReplyDispatcherOptions = {
   deliver: ReplyDispatchDeliverer;
   responsePrefix?: string;
+  /** Static context for response prefix template interpolation. */
+  responsePrefixContext?: ResponsePrefixContext;
+  /** Dynamic context provider for response prefix template interpolation.
+   * Called at normalization time, after model selection is complete. */
+  responsePrefixContextProvider?: () => ResponsePrefixContext;
   onHeartbeatStrip?: () => void;
   onIdle?: () => void;
   onError?: ReplyDispatchErrorHandler;
@@ -61,10 +67,17 @@ export type ReplyDispatcher = {
 
 function normalizeReplyPayloadInternal(
   payload: ReplyPayload,
-  opts: Pick<ReplyDispatcherOptions, "responsePrefix" | "onHeartbeatStrip">,
+  opts: Pick<
+    ReplyDispatcherOptions,
+    "responsePrefix" | "responsePrefixContext" | "responsePrefixContextProvider" | "onHeartbeatStrip"
+  >,
 ): ReplyPayload | null {
+  // Prefer dynamic context provider over static context
+  const prefixContext = opts.responsePrefixContextProvider?.() ?? opts.responsePrefixContext;
+
   return normalizeReplyPayload(payload, {
     responsePrefix: opts.responsePrefix,
+    responsePrefixContext: prefixContext,
     onHeartbeatStrip: opts.onHeartbeatStrip,
   });
 }
diff --git a/src/auto-reply/reply/response-prefix-template.test.ts b/src/auto-reply/reply/response-prefix-template.test.ts
new file mode 100644
index 000000000..db899f44d
--- /dev/null
+++ b/src/auto-reply/reply/response-prefix-template.test.ts
@@ -0,0 +1,181 @@
+import { describe, expect, it } from "vitest";
+
+import {
+  extractShortModelName,
+  hasTemplateVariables,
+  resolveResponsePrefixTemplate,
+} from "./response-prefix-template.js";
+
+describe("resolveResponsePrefixTemplate", () => {
+  it("returns undefined for undefined template", () => {
+    expect(resolveResponsePrefixTemplate(undefined, {})).toBeUndefined();
+  });
+
+  it("returns template as-is when no variables present", () => {
+    expect(resolveResponsePrefixTemplate("[Claude]", {})).toBe("[Claude]");
+  });
+
+  it("resolves {model} variable", () => {
+    const result = resolveResponsePrefixTemplate("[{model}]", {
+      model: "gpt-5.2",
+    });
+    expect(result).toBe("[gpt-5.2]");
+  });
+
+  it("resolves {modelFull} variable", () => {
+    const result = resolveResponsePrefixTemplate("[{modelFull}]", {
+      modelFull: "openai-codex/gpt-5.2",
+    });
+    expect(result).toBe("[openai-codex/gpt-5.2]");
+  });
+
+  it("resolves {provider} variable", () => {
+    const result = resolveResponsePrefixTemplate("[{provider}]", {
+      provider: "anthropic",
+    });
+    expect(result).toBe("[anthropic]");
+  });
+
+  it("resolves {thinkingLevel} variable", () => {
+    const result = resolveResponsePrefixTemplate("think:{thinkingLevel}", {
+      thinkingLevel: "high",
+    });
+    expect(result).toBe("think:high");
+  });
+
+  it("resolves {think} as alias for thinkingLevel", () => {
+    const result = resolveResponsePrefixTemplate("think:{think}", {
+      thinkingLevel: "low",
+    });
+    expect(result).toBe("think:low");
+  });
+
+  it("resolves {identity.name} variable", () => {
+    const result = resolveResponsePrefixTemplate("[{identity.name}]", {
+      identityName: "Clawdbot",
+    });
+    expect(result).toBe("[Clawdbot]");
+  });
+
+  it("resolves {identityName} as alias", () => {
+    const result = resolveResponsePrefixTemplate("[{identityName}]", {
+      identityName: "Clawdbot",
+    });
+    expect(result).toBe("[Clawdbot]");
+  });
+
+  it("resolves multiple variables", () => {
+    const result = resolveResponsePrefixTemplate("[{model} | think:{thinkingLevel}]", {
+      model: "claude-opus-4-5",
+      thinkingLevel: "high",
+    });
+    expect(result).toBe("[claude-opus-4-5 | think:high]");
+  });
+
+  it("leaves unresolved variables as-is", () => {
+    const result = resolveResponsePrefixTemplate("[{model}]", {});
+    expect(result).toBe("[{model}]");
+  });
+
+  it("leaves unrecognized variables as-is", () => {
+    const result = resolveResponsePrefixTemplate("[{unknownVar}]", {
+      model: "gpt-5.2",
+    });
+    expect(result).toBe("[{unknownVar}]");
+  });
+
+  it("handles case insensitivity", () => {
+    const result = resolveResponsePrefixTemplate("[{MODEL} | {ThinkingLevel}]", {
+      model: "gpt-5.2",
+      thinkingLevel: "low",
+    });
+    expect(result).toBe("[gpt-5.2 | low]");
+  });
+
+  it("handles mixed resolved and unresolved variables", () => {
+    const result = resolveResponsePrefixTemplate("[{model} | {provider}]", {
+      model: "gpt-5.2",
+      // provider not provided
+    });
+    expect(result).toBe("[gpt-5.2 | {provider}]");
+  });
+
+  it("handles complex template with all variables", () => {
+    const result = resolveResponsePrefixTemplate(
+      "[{identity.name}] {provider}/{model} (think:{thinkingLevel})",
+      {
+        identityName: "Clawdbot",
+        provider: "anthropic",
+        model: "claude-opus-4-5",
+        thinkingLevel: "high",
+      },
+    );
+    expect(result).toBe("[Clawdbot] anthropic/claude-opus-4-5 (think:high)");
+  });
+});
+
+describe("extractShortModelName", () => {
+  it("strips provider prefix", () => {
+    expect(extractShortModelName("openai/gpt-5.2")).toBe("gpt-5.2");
+    expect(extractShortModelName("anthropic/claude-opus-4-5")).toBe("claude-opus-4-5");
+    expect(extractShortModelName("openai-codex/gpt-5.2-codex")).toBe("gpt-5.2-codex");
+  });
+
+  it("strips date suffix", () => {
+    expect(extractShortModelName("claude-opus-4-5-20251101")).toBe("claude-opus-4-5");
+    expect(extractShortModelName("gpt-5.2-20250115")).toBe("gpt-5.2");
+  });
+
+  it("strips -latest suffix", () => {
+    expect(extractShortModelName("gpt-5.2-latest")).toBe("gpt-5.2");
+    expect(extractShortModelName("claude-sonnet-latest")).toBe("claude-sonnet");
+  });
+
+  it("handles model without provider", () => {
+    expect(extractShortModelName("gpt-5.2")).toBe("gpt-5.2");
+    expect(extractShortModelName("claude-opus-4-5")).toBe("claude-opus-4-5");
+  });
+
+  it("handles full path with provider and date suffix", () => {
+    expect(extractShortModelName("anthropic/claude-opus-4-5-20251101")).toBe("claude-opus-4-5");
+  });
+
+  it("preserves version numbers that look like dates but are not", () => {
+    // Date suffix must be exactly 8 digits at the end
+    expect(extractShortModelName("model-v1234567")).toBe("model-v1234567");
+    expect(extractShortModelName("model-123456789")).toBe("model-123456789");
+  });
+});
+
+describe("hasTemplateVariables", () => {
+  it("returns false for undefined", () => {
+    expect(hasTemplateVariables(undefined)).toBe(false);
+  });
+
+  it("returns false for empty string", () => {
+    expect(hasTemplateVariables("")).toBe(false);
+  });
+
+  it("returns false for static prefix", () => {
+    expect(hasTemplateVariables("[Claude]")).toBe(false);
+  });
+
+  it("returns true when template variables present", () => {
+    expect(hasTemplateVariables("[{model}]")).toBe(true);
+    expect(hasTemplateVariables("{provider}")).toBe(true);
+    expect(hasTemplateVariables("prefix {thinkingLevel} suffix")).toBe(true);
+  });
+
+  it("returns true for multiple variables", () => {
+    expect(hasTemplateVariables("[{model} | {provider}]")).toBe(true);
+  });
+
+  it("handles consecutive calls correctly (regex lastIndex reset)", () => {
+    // First call
+    expect(hasTemplateVariables("[{model}]")).toBe(true);
+    // Second call should still work
+    expect(hasTemplateVariables("[{model}]")).toBe(true);
+    // Static string should return false
+    expect(hasTemplateVariables("[Claude]")).toBe(false);
+  });
+});
diff --git a/src/auto-reply/reply/response-prefix-template.ts b/src/auto-reply/reply/response-prefix-template.ts
new file mode 100644
index 000000000..788531ca5
--- /dev/null
+++ b/src/auto-reply/reply/response-prefix-template.ts
@@ -0,0 +1,97 @@
+/**
+ * Template interpolation for response prefix.
+ *
+ * Supports variables like `{model}`, `{provider}`, `{thinkingLevel}`, etc.
+ * Variables are case-insensitive and unresolved ones remain as literal text.
+ */
+
+export type ResponsePrefixContext = {
+  /** Short model name (e.g., "gpt-5.2", "claude-opus-4-5") */
+  model?: string;
+  /** Full model ID including provider (e.g., "openai-codex/gpt-5.2") */
+  modelFull?: string;
+  /** Provider name (e.g., "openai-codex", "anthropic") */
+  provider?: string;
+  /** Current thinking level (e.g., "high", "low", "off") */
+  thinkingLevel?: string;
+  /** Agent identity name */
+  identityName?: string;
+};
+
+// Regex pattern for template variables: {variableName} or {variable.name}
+const TEMPLATE_VAR_PATTERN = /\{([a-zA-Z][a-zA-Z0-9.]*)\}/g;
+
+/**
+ * Interpolate template variables in a response prefix string.
+ *
+ * @param template - The template string with `{variable}` placeholders
+ * @param context - Context object with values for interpolation
+ * @returns The interpolated string, or undefined if template is undefined
+ *
+ * @example
+ * resolveResponsePrefixTemplate("[{model} | think:{thinkingLevel}]", {
+ *   model: "gpt-5.2",
+ *   thinkingLevel: "high"
+ * })
+ * // Returns: "[gpt-5.2 | think:high]"
+ */
+export function resolveResponsePrefixTemplate(
+  template: string | undefined,
+  context: ResponsePrefixContext,
+): string | undefined {
+  if (!template) return undefined;
+
+  return template.replace(TEMPLATE_VAR_PATTERN, (match, varName: string) => {
+    const normalizedVar = varName.toLowerCase();
+
+    switch (normalizedVar) {
+      case "model":
+        return context.model ?? match;
+      case "modelfull":
+        return context.modelFull ?? match;
+      case "provider":
+        return context.provider ?? match;
+      case "thinkinglevel":
+      case "think":
+        return context.thinkingLevel ?? match;
+      case "identity.name":
+      case "identityname":
+        return context.identityName ?? match;
+      default:
+        // Leave unrecognized variables as-is
+        return match;
+    }
+  });
+}
+
+/**
+ * Extract short model name from a full model string.
+ *
+ * Strips:
+ * - Provider prefix (e.g., "openai/" from "openai/gpt-5.2")
+ * - Date suffixes (e.g., "-20251101" from "claude-opus-4-5-20251101")
+ * - Common version suffixes (e.g., "-latest")
+ *
+ * @example
+ * extractShortModelName("openai-codex/gpt-5.2") // "gpt-5.2"
+ * extractShortModelName("claude-opus-4-5-20251101") // "claude-opus-4-5"
+ * extractShortModelName("gpt-5.2-latest") // "gpt-5.2"
+ */
+export function extractShortModelName(fullModel: string): string {
+  // Strip provider prefix
+  const slash = fullModel.lastIndexOf("/");
+  const modelPart = slash >= 0 ? fullModel.slice(slash + 1) : fullModel;
+
+  // Strip date suffixes (YYYYMMDD format)
+  return modelPart.replace(/-\d{8}$/, "").replace(/-latest$/, "");
+}
+
+/**
+ * Check if a template string contains any template variables.
+ */
+export function hasTemplateVariables(template: string | undefined): boolean {
+  if (!template) return false;
+  // Reset lastIndex since we're using a global regex
+  TEMPLATE_VAR_PATTERN.lastIndex = 0;
+  return TEMPLATE_VAR_PATTERN.test(template);
+}
diff --git a/src/auto-reply/types.ts b/src/auto-reply/types.ts
index bf6e765e1..e1bf611db 100644
--- a/src/auto-reply/types.ts
+++ b/src/auto-reply/types.ts
@@ -5,6 +5,13 @@ export type BlockReplyContext = {
   timeoutMs?: number;
 };
 
+/** Context passed to onModelSelected callback with actual model used. */
+export type ModelSelectedContext = {
+  provider: string;
+  model: string;
+  thinkLevel: string | undefined;
+};
+
 export type GetReplyOptions = {
   onReplyStart?: () => Promise<void> | void;
   onTypingController?: (typing: TypingController) => void;
@@ -13,6 +20,9 @@ export type GetReplyOptions = {
   onReasoningStream?: (payload: ReplyPayload) => Promise<void> | void;
   onBlockReply?: (payload: ReplyPayload, context?: BlockReplyContext) => Promise<void> | void;
   onToolResult?: (payload: ReplyPayload) => Promise<void> | void;
+  /** Called when the actual model is selected (including after fallback).
+   * Use this to get model/provider/thinkLevel for responsePrefix template interpolation. */
+  onModelSelected?: (ctx: ModelSelectedContext) => void;
   disableBlockStreaming?: boolean;
   /** Timeout for block reply delivery (ms). */
   blockReplyTimeoutMs?: number;
diff --git a/src/config/types.messages.ts b/src/config/types.messages.ts
index 76754cb80..d137b56d8 100644
--- a/src/config/types.messages.ts
+++ b/src/config/types.messages.ts
@@ -44,8 +44,21 @@ export type MessagesConfig = {
   messagePrefix?: string;
   /**
    * Prefix auto-added to all outbound replies.
-   * - string: explicit prefix
+   *
+   * - string: explicit prefix (may include template variables)
    * - special value: `"auto"` derives `[{agents.list[].identity.name}]` for the routed agent (when set)
+   *
+   * Supported template variables (case-insensitive):
+   * - `{model}` - short model name (e.g., `claude-opus-4-5`, `gpt-4o`)
+   * - `{modelFull}` - full model identifier (e.g., `anthropic/claude-opus-4-5`)
+   * - `{provider}` - provider name (e.g., `anthropic`, `openai`)
+   * - `{thinkingLevel}` or `{think}` - current thinking level (`high`, `low`, `off`)
+   * - `{identity.name}` or `{identityName}` - agent identity name
+   *
+   * Example: `"[{model} | think:{thinkingLevel}]"` → `"[claude-opus-4-5 | think:high]"`
+   *
+   * Unresolved variables remain as literal text (e.g., `{model}` if context unavailable).
+   *
    * Default: none
    */
   responsePrefix?: string;
diff --git a/src/discord/monitor/message-handler.process.ts b/src/discord/monitor/message-handler.process.ts
index b4862edba..4758feb09 100644
--- a/src/discord/monitor/message-handler.process.ts
+++ b/src/discord/monitor/message-handler.process.ts
@@ -2,7 +2,12 @@ import {
   resolveAckReaction,
   resolveEffectiveMessagesConfig,
   resolveHumanDelayConfig,
+  resolveIdentityName,
 } from "../../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../auto-reply/reply/response-prefix-template.js";
 import { formatAgentEnvelope, formatThreadStarterEnvelope } from "../../auto-reply/envelope.js";
 import { dispatchReplyFromConfig } from "../../auto-reply/reply/dispatch-from-config.js";
 import { buildHistoryContextFromMap, clearHistoryEntries } from "../../auto-reply/reply/history.js";
@@ -280,8 +285,15 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
   const typingChannelId = deliverTarget.startsWith("channel:")
     ? deliverTarget.slice("channel:".length)
     : message.channelId;
+
+  // Create mutable context for response prefix template interpolation
+  let prefixContext: ResponsePrefixContext = {
+    identityName: resolveIdentityName(cfg, route.agentId),
+  };
+
   const { dispatcher, replyOptions, markDispatchIdle } = createReplyDispatcherWithTyping({
     responsePrefix: resolveEffectiveMessagesConfig(cfg, route.agentId).responsePrefix,
+    responsePrefixContextProvider: () => prefixContext,
     humanDelay: resolveHumanDelayConfig(cfg, route.agentId),
     deliver: async (payload: ReplyPayload) => {
       const replyToId = replyReference.use();
@@ -316,6 +328,15 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
         typeof discordConfig?.blockStreaming === "boolean"
           ? !discordConfig.blockStreaming
           : undefined,
+      onModelSelected: (ctx) => {
+        prefixContext = {
+          ...prefixContext,
+          provider: ctx.provider,
+          model: extractShortModelName(ctx.model),
+          modelFull: `${ctx.provider}/${ctx.model}`,
+          thinkingLevel: ctx.thinkLevel ?? "off",
+        };
+      },
     },
   });
   markDispatchIdle();
diff --git a/src/imessage/monitor/monitor-provider.ts b/src/imessage/monitor/monitor-provider.ts
index 831219c04..b8a408419 100644
--- a/src/imessage/monitor/monitor-provider.ts
+++ b/src/imessage/monitor/monitor-provider.ts
@@ -1,4 +1,12 @@
-import { resolveEffectiveMessagesConfig, resolveHumanDelayConfig } from "../../agents/identity.js";
+import {
+  resolveEffectiveMessagesConfig,
+  resolveHumanDelayConfig,
+  resolveIdentityName,
+} from "../../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../auto-reply/reply/response-prefix-template.js";
 import { resolveTextChunkLimit } from "../../auto-reply/chunk.js";
 import { hasControlCommand } from "../../auto-reply/command-detection.js";
 import { formatAgentEnvelope } from "../../auto-reply/envelope.js";
@@ -341,8 +349,15 @@ export async function monitorIMessageProvider(opts: MonitorIMessageOpts = {}): P
     }
 
     let didSendReply = false;
+
+    // Create mutable context for response prefix template interpolation
+    let prefixContext: ResponsePrefixContext = {
+      identityName: resolveIdentityName(cfg, route.agentId),
+    };
+
     const dispatcher = createReplyDispatcher({
       responsePrefix: resolveEffectiveMessagesConfig(cfg, route.agentId).responsePrefix,
+      responsePrefixContextProvider: () => prefixContext,
       humanDelay: resolveHumanDelayConfig(cfg, route.agentId),
       deliver: async (payload) => {
         await deliverReplies({
@@ -370,6 +385,15 @@ export async function monitorIMessageProvider(opts: MonitorIMessageOpts = {}): P
           typeof accountInfo.config.blockStreaming === "boolean"
             ? !accountInfo.config.blockStreaming
             : undefined,
+        onModelSelected: (ctx) => {
+          prefixContext = {
+            ...prefixContext,
+            provider: ctx.provider,
+            model: extractShortModelName(ctx.model),
+            modelFull: `${ctx.provider}/${ctx.model}`,
+            thinkingLevel: ctx.thinkLevel ?? "off",
+          };
+        },
       },
     });
     if (!queuedFinal) {
diff --git a/src/signal/monitor/event-handler.ts b/src/signal/monitor/event-handler.ts
index bcc77ec9d..7f093a3c1 100644
--- a/src/signal/monitor/event-handler.ts
+++ b/src/signal/monitor/event-handler.ts
@@ -1,4 +1,12 @@
-import { resolveEffectiveMessagesConfig, resolveHumanDelayConfig } from "../../agents/identity.js";
+import {
+  resolveEffectiveMessagesConfig,
+  resolveHumanDelayConfig,
+  resolveIdentityName,
+} from "../../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../auto-reply/reply/response-prefix-template.js";
 import { formatAgentEnvelope } from "../../auto-reply/envelope.js";
 import { dispatchReplyFromConfig } from "../../auto-reply/reply/dispatch-from-config.js";
 import { buildHistoryContextFromMap, clearHistoryEntries } from "../../auto-reply/reply/history.js";
@@ -310,8 +318,15 @@ export function createSignalEventHandler(deps: SignalEventHandlerDeps) {
     }
 
     let didSendReply = false;
+
+    // Create mutable context for response prefix template interpolation
+    let prefixContext: ResponsePrefixContext = {
+      identityName: resolveIdentityName(deps.cfg, route.agentId),
+    };
+
     const dispatcher = createReplyDispatcher({
       responsePrefix: resolveEffectiveMessagesConfig(deps.cfg, route.agentId).responsePrefix,
+      responsePrefixContextProvider: () => prefixContext,
       humanDelay: resolveHumanDelayConfig(deps.cfg, route.agentId),
       deliver: async (payload) => {
         await deps.deliverReplies({
@@ -338,6 +353,15 @@ export function createSignalEventHandler(deps: SignalEventHandlerDeps) {
       replyOptions: {
         disableBlockStreaming:
           typeof deps.blockStreaming === "boolean" ? !deps.blockStreaming : undefined,
+        onModelSelected: (ctx) => {
+          prefixContext = {
+            ...prefixContext,
+            provider: ctx.provider,
+            model: extractShortModelName(ctx.model),
+            modelFull: `${ctx.provider}/${ctx.model}`,
+            thinkingLevel: ctx.thinkLevel ?? "off",
+          };
+        },
       },
     });
     if (!queuedFinal) {
diff --git a/src/slack/monitor/message-handler/dispatch.ts b/src/slack/monitor/message-handler/dispatch.ts
index dc740d43c..836e85d46 100644
--- a/src/slack/monitor/message-handler/dispatch.ts
+++ b/src/slack/monitor/message-handler/dispatch.ts
@@ -1,7 +1,12 @@
 import {
   resolveEffectiveMessagesConfig,
   resolveHumanDelayConfig,
+  resolveIdentityName,
 } from "../../../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../../auto-reply/reply/response-prefix-template.js";
 import { dispatchReplyFromConfig } from "../../../auto-reply/reply/dispatch-from-config.js";
 import { clearHistoryEntries } from "../../../auto-reply/reply/history.js";
 import { createReplyDispatcherWithTyping } from "../../../auto-reply/reply/reply-dispatcher.js";
@@ -62,8 +67,15 @@ export async function dispatchPreparedSlackMessage(prepared: PreparedSlackMessag
   };
 
   let didSendReply = false;
+
+  // Create mutable context for response prefix template interpolation
+  let prefixContext: ResponsePrefixContext = {
+    identityName: resolveIdentityName(cfg, route.agentId),
+  };
+
   const { dispatcher, replyOptions, markDispatchIdle } = createReplyDispatcherWithTyping({
     responsePrefix: resolveEffectiveMessagesConfig(cfg, route.agentId).responsePrefix,
+    responsePrefixContextProvider: () => prefixContext,
     humanDelay: resolveHumanDelayConfig(cfg, route.agentId),
     deliver: async (payload) => {
       const replyThreadTs = replyPlan.nextThreadTs();
@@ -104,6 +116,15 @@ export async function dispatchPreparedSlackMessage(prepared: PreparedSlackMessag
         typeof account.config.blockStreaming === "boolean"
           ? !account.config.blockStreaming
           : undefined,
+      onModelSelected: (ctx) => {
+        prefixContext = {
+          ...prefixContext,
+          provider: ctx.provider,
+          model: extractShortModelName(ctx.model),
+          modelFull: `${ctx.provider}/${ctx.model}`,
+          thinkingLevel: ctx.thinkLevel ?? "off",
+        };
+      },
     },
   });
   markDispatchIdle();
diff --git a/src/telegram/bot-message-dispatch.ts b/src/telegram/bot-message-dispatch.ts
index 195f2872c..bae515f4b 100644
--- a/src/telegram/bot-message-dispatch.ts
+++ b/src/telegram/bot-message-dispatch.ts
@@ -1,5 +1,9 @@
 // @ts-nocheck
-import { resolveEffectiveMessagesConfig } from "../agents/identity.js";
+import { resolveEffectiveMessagesConfig, resolveIdentityName } from "../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../auto-reply/reply/response-prefix-template.js";
 import { EmbeddedBlockChunker } from "../agents/pi-embedded-block-chunker.js";
 import { clearHistoryEntries } from "../auto-reply/reply/history.js";
 import { dispatchReplyWithBufferedBlockDispatcher } from "../auto-reply/reply/provider-dispatcher.js";
@@ -114,12 +118,18 @@ export const dispatchTelegramMessage = async ({
     Boolean(draftStream) ||
     (typeof telegramCfg.blockStreaming === "boolean" ? !telegramCfg.blockStreaming : undefined);
 
+  // Create mutable context for response prefix template interpolation
+  let prefixContext: ResponsePrefixContext = {
+    identityName: resolveIdentityName(cfg, route.agentId),
+  };
+
   let didSendReply = false;
   const { queuedFinal } = await dispatchReplyWithBufferedBlockDispatcher({
     ctx: ctxPayload,
     cfg,
     dispatcherOptions: {
       responsePrefix: resolveEffectiveMessagesConfig(cfg, route.agentId).responsePrefix,
+      responsePrefixContextProvider: () => prefixContext,
       deliver: async (payload, info) => {
         if (info.kind === "final") {
           await flushDraft();
@@ -151,6 +161,15 @@ export const dispatchTelegramMessage = async ({
           }
         : undefined,
       disableBlockStreaming,
+      onModelSelected: (ctx) => {
+        prefixContext = {
+          ...prefixContext,
+          provider: ctx.provider,
+          model: extractShortModelName(ctx.model),
+          modelFull: `${ctx.provider}/${ctx.model}`,
+          thinkingLevel: ctx.thinkLevel ?? "off",
+        };
+      },
     },
   });
   draftStream?.stop();
diff --git a/src/web/auto-reply/monitor/process-message.ts b/src/web/auto-reply/monitor/process-message.ts
index 4ae3b3fc0..0a57e6877 100644
--- a/src/web/auto-reply/monitor/process-message.ts
+++ b/src/web/auto-reply/monitor/process-message.ts
@@ -1,4 +1,8 @@
-import { resolveEffectiveMessagesConfig } from "../../../agents/identity.js";
+import { resolveEffectiveMessagesConfig, resolveIdentityName } from "../../../agents/identity.js";
+import {
+  extractShortModelName,
+  type ResponsePrefixContext,
+} from "../../../auto-reply/reply/response-prefix-template.js";
 import { resolveTextChunkLimit } from "../../../auto-reply/chunk.js";
 import { formatAgentEnvelope } from "../../../auto-reply/envelope.js";
 import { buildHistoryContext } from "../../../auto-reply/reply/history.js";
@@ -173,6 +177,11 @@ export async function processMessage(params: {
     params.route.agentId,
   ).responsePrefix;
 
+  // Create mutable context for response prefix template interpolation
+  let prefixContext: ResponsePrefixContext = {
+    identityName: resolveIdentityName(params.cfg, params.route.agentId),
+  };
+
   const { queuedFinal } = await dispatchReplyWithBufferedBlockDispatcher({
     ctx: {
       Body: combinedBody,
@@ -210,6 +219,7 @@ export async function processMessage(params: {
     replyResolver: params.replyResolver,
     dispatcherOptions: {
       responsePrefix,
+      responsePrefixContextProvider: () => prefixContext,
       onHeartbeatStrip: () => {
         if (!didLogHeartbeatStrip) {
           didLogHeartbeatStrip = true;
@@ -267,6 +277,15 @@ export async function processMessage(params: {
         typeof params.cfg.channels?.whatsapp?.blockStreaming === "boolean"
           ? !params.cfg.channels.whatsapp.blockStreaming
           : undefined,
+      onModelSelected: (ctx) => {
+        prefixContext = {
+          ...prefixContext,
+          provider: ctx.provider,
+          model: extractShortModelName(ctx.model),
+          modelFull: `${ctx.provider}/${ctx.model}`,
+          thinkingLevel: ctx.thinkLevel ?? "off",
+        };
+      },
     },
   });
 

From 7b04e6ac42ac62963ce033325a652e1bfc65689e Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:15:46 -0500
Subject: [PATCH 2/6] debug: add prefix template resolution logging

---
 src/auto-reply/reply/agent-runner-execution.ts | 3 +++
 src/auto-reply/reply/reply-dispatcher.ts       | 8 ++++++++
 2 files changed, 11 insertions(+)

diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index d1f0e276d..b580dcc4b 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -127,6 +127,9 @@ export async function runAgentTurnWithFallback(params: {
         run: (provider, model) => {
           // Notify that model selection is complete (including after fallback).
           // This allows responsePrefix template interpolation with the actual model.
+          logVerbose(
+            `[prefix-debug] onModelSelected firing: provider=${provider} model=${model} thinkLevel=${params.followupRun.run.thinkLevel}`,
+          );
           params.opts?.onModelSelected?.({
             provider,
             model,
diff --git a/src/auto-reply/reply/reply-dispatcher.ts b/src/auto-reply/reply/reply-dispatcher.ts
index b8eeca477..4a45161d8 100644
--- a/src/auto-reply/reply/reply-dispatcher.ts
+++ b/src/auto-reply/reply/reply-dispatcher.ts
@@ -75,6 +75,14 @@ function normalizeReplyPayloadInternal(
   // Prefer dynamic context provider over static context
   const prefixContext = opts.responsePrefixContextProvider?.() ?? opts.responsePrefixContext;
 
+  // Debug logging for prefix template resolution
+  if (opts.responsePrefix?.includes("{")) {
+    // eslint-disable-next-line no-console
+    console.log(
+      `[prefix-debug] normalizing with context: ${JSON.stringify(prefixContext)} prefix: ${opts.responsePrefix}`,
+    );
+  }
+
   return normalizeReplyPayload(payload, {
     responsePrefix: opts.responsePrefix,
     responsePrefixContext: prefixContext,

From 113eea5047c49e11b5d014fafcb9c2c7bab5c2c2 Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:20:19 -0500
Subject: [PATCH 3/6] fix: mutate prefixContext object instead of reassigning
 for closure correctness

---
 src/auto-reply/reply/agent-runner-execution.ts |  3 ---
 src/auto-reply/reply/reply-dispatcher.ts       |  8 --------
 src/discord/monitor/message-handler.process.ts | 12 +++++-------
 src/imessage/monitor/monitor-provider.ts       | 12 +++++-------
 src/signal/monitor/event-handler.ts            | 12 +++++-------
 src/slack/monitor/message-handler/dispatch.ts  | 12 +++++-------
 src/telegram/bot-message-dispatch.ts           | 12 +++++-------
 src/web/auto-reply/monitor/process-message.ts  | 12 +++++-------
 8 files changed, 30 insertions(+), 53 deletions(-)

diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index b580dcc4b..d1f0e276d 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -127,9 +127,6 @@ export async function runAgentTurnWithFallback(params: {
         run: (provider, model) => {
           // Notify that model selection is complete (including after fallback).
           // This allows responsePrefix template interpolation with the actual model.
-          logVerbose(
-            `[prefix-debug] onModelSelected firing: provider=${provider} model=${model} thinkLevel=${params.followupRun.run.thinkLevel}`,
-          );
           params.opts?.onModelSelected?.({
             provider,
             model,
diff --git a/src/auto-reply/reply/reply-dispatcher.ts b/src/auto-reply/reply/reply-dispatcher.ts
index 4a45161d8..b8eeca477 100644
--- a/src/auto-reply/reply/reply-dispatcher.ts
+++ b/src/auto-reply/reply/reply-dispatcher.ts
@@ -75,14 +75,6 @@ function normalizeReplyPayloadInternal(
   // Prefer dynamic context provider over static context
   const prefixContext = opts.responsePrefixContextProvider?.() ?? opts.responsePrefixContext;
 
-  // Debug logging for prefix template resolution
-  if (opts.responsePrefix?.includes("{")) {
-    // eslint-disable-next-line no-console
-    console.log(
-      `[prefix-debug] normalizing with context: ${JSON.stringify(prefixContext)} prefix: ${opts.responsePrefix}`,
-    );
-  }
-
   return normalizeReplyPayload(payload, {
     responsePrefix: opts.responsePrefix,
     responsePrefixContext: prefixContext,
diff --git a/src/discord/monitor/message-handler.process.ts b/src/discord/monitor/message-handler.process.ts
index 4758feb09..dae301d04 100644
--- a/src/discord/monitor/message-handler.process.ts
+++ b/src/discord/monitor/message-handler.process.ts
@@ -329,13 +329,11 @@ export async function processDiscordMessage(ctx: DiscordMessagePreflightContext)
           ? !discordConfig.blockStreaming
           : undefined,
       onModelSelected: (ctx) => {
-        prefixContext = {
-          ...prefixContext,
-          provider: ctx.provider,
-          model: extractShortModelName(ctx.model),
-          modelFull: `${ctx.provider}/${ctx.model}`,
-          thinkingLevel: ctx.thinkLevel ?? "off",
-        };
+        // Mutate the object directly instead of reassigning to ensure the closure sees updates
+        prefixContext.provider = ctx.provider;
+        prefixContext.model = extractShortModelName(ctx.model);
+        prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+        prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
       },
     },
   });
diff --git a/src/imessage/monitor/monitor-provider.ts b/src/imessage/monitor/monitor-provider.ts
index b8a408419..1d86ec993 100644
--- a/src/imessage/monitor/monitor-provider.ts
+++ b/src/imessage/monitor/monitor-provider.ts
@@ -386,13 +386,11 @@ export async function monitorIMessageProvider(opts: MonitorIMessageOpts = {}): P
             ? !accountInfo.config.blockStreaming
             : undefined,
         onModelSelected: (ctx) => {
-          prefixContext = {
-            ...prefixContext,
-            provider: ctx.provider,
-            model: extractShortModelName(ctx.model),
-            modelFull: `${ctx.provider}/${ctx.model}`,
-            thinkingLevel: ctx.thinkLevel ?? "off",
-          };
+          // Mutate the object directly instead of reassigning to ensure the closure sees updates
+          prefixContext.provider = ctx.provider;
+          prefixContext.model = extractShortModelName(ctx.model);
+          prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+          prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
         },
       },
     });
diff --git a/src/signal/monitor/event-handler.ts b/src/signal/monitor/event-handler.ts
index 7f093a3c1..0aac7c97b 100644
--- a/src/signal/monitor/event-handler.ts
+++ b/src/signal/monitor/event-handler.ts
@@ -354,13 +354,11 @@ export function createSignalEventHandler(deps: SignalEventHandlerDeps) {
         disableBlockStreaming:
           typeof deps.blockStreaming === "boolean" ? !deps.blockStreaming : undefined,
         onModelSelected: (ctx) => {
-          prefixContext = {
-            ...prefixContext,
-            provider: ctx.provider,
-            model: extractShortModelName(ctx.model),
-            modelFull: `${ctx.provider}/${ctx.model}`,
-            thinkingLevel: ctx.thinkLevel ?? "off",
-          };
+          // Mutate the object directly instead of reassigning to ensure the closure sees updates
+          prefixContext.provider = ctx.provider;
+          prefixContext.model = extractShortModelName(ctx.model);
+          prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+          prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
         },
       },
     });
diff --git a/src/slack/monitor/message-handler/dispatch.ts b/src/slack/monitor/message-handler/dispatch.ts
index 836e85d46..fb0eb66b1 100644
--- a/src/slack/monitor/message-handler/dispatch.ts
+++ b/src/slack/monitor/message-handler/dispatch.ts
@@ -117,13 +117,11 @@ export async function dispatchPreparedSlackMessage(prepared: PreparedSlackMessag
           ? !account.config.blockStreaming
           : undefined,
       onModelSelected: (ctx) => {
-        prefixContext = {
-          ...prefixContext,
-          provider: ctx.provider,
-          model: extractShortModelName(ctx.model),
-          modelFull: `${ctx.provider}/${ctx.model}`,
-          thinkingLevel: ctx.thinkLevel ?? "off",
-        };
+        // Mutate the object directly instead of reassigning to ensure the closure sees updates
+        prefixContext.provider = ctx.provider;
+        prefixContext.model = extractShortModelName(ctx.model);
+        prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+        prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
       },
     },
   });
diff --git a/src/telegram/bot-message-dispatch.ts b/src/telegram/bot-message-dispatch.ts
index bae515f4b..c609a07c2 100644
--- a/src/telegram/bot-message-dispatch.ts
+++ b/src/telegram/bot-message-dispatch.ts
@@ -162,13 +162,11 @@ export const dispatchTelegramMessage = async ({
         : undefined,
       disableBlockStreaming,
       onModelSelected: (ctx) => {
-        prefixContext = {
-          ...prefixContext,
-          provider: ctx.provider,
-          model: extractShortModelName(ctx.model),
-          modelFull: `${ctx.provider}/${ctx.model}`,
-          thinkingLevel: ctx.thinkLevel ?? "off",
-        };
+        // Mutate the object directly instead of reassigning to ensure the closure sees updates
+        prefixContext.provider = ctx.provider;
+        prefixContext.model = extractShortModelName(ctx.model);
+        prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+        prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
       },
     },
   });
diff --git a/src/web/auto-reply/monitor/process-message.ts b/src/web/auto-reply/monitor/process-message.ts
index 0a57e6877..205fd5699 100644
--- a/src/web/auto-reply/monitor/process-message.ts
+++ b/src/web/auto-reply/monitor/process-message.ts
@@ -278,13 +278,11 @@ export async function processMessage(params: {
           ? !params.cfg.channels.whatsapp.blockStreaming
           : undefined,
       onModelSelected: (ctx) => {
-        prefixContext = {
-          ...prefixContext,
-          provider: ctx.provider,
-          model: extractShortModelName(ctx.model),
-          modelFull: `${ctx.provider}/${ctx.model}`,
-          thinkingLevel: ctx.thinkLevel ?? "off",
-        };
+        // Mutate the object directly instead of reassigning to ensure the closure sees updates
+        prefixContext.provider = ctx.provider;
+        prefixContext.model = extractShortModelName(ctx.model);
+        prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
+        prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
       },
     },
   });

From 56b3b44342d79793ce34a9caf0ca1ab5069f0863 Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:23:21 -0500
Subject: [PATCH 4/6] debug: add responsePrefix template logging

---
 src/auto-reply/reply/agent-runner-execution.ts | 3 +++
 src/auto-reply/reply/normalize-reply.ts        | 9 +++++++++
 src/telegram/bot-message-dispatch.ts           | 4 ++++
 3 files changed, 16 insertions(+)

diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index d1f0e276d..49a50cd29 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -127,6 +127,9 @@ export async function runAgentTurnWithFallback(params: {
         run: (provider, model) => {
           // Notify that model selection is complete (including after fallback).
           // This allows responsePrefix template interpolation with the actual model.
+          logVerbose(
+            `[responsePrefix] onModelSelected callback exists: ${!!params.opts?.onModelSelected}, provider=${provider}, model=${model}`,
+          );
           params.opts?.onModelSelected?.({
             provider,
             model,
diff --git a/src/auto-reply/reply/normalize-reply.ts b/src/auto-reply/reply/normalize-reply.ts
index ac8e85f2e..a9d0d0060 100644
--- a/src/auto-reply/reply/normalize-reply.ts
+++ b/src/auto-reply/reply/normalize-reply.ts
@@ -1,3 +1,4 @@
+import { logVerbose } from "../../globals.js";
 import { stripHeartbeatToken } from "../heartbeat.js";
 import { HEARTBEAT_TOKEN, isSilentReplyText, SILENT_REPLY_TOKEN } from "../tokens.js";
 import type { ReplyPayload } from "../types.js";
@@ -43,9 +44,17 @@ export function normalizeReplyPayload(
   }
 
   // Resolve template variables in responsePrefix if context is provided
+  if (opts.responsePrefix?.includes("{")) {
+    logVerbose(
+      `[responsePrefix] normalizing: prefix="${opts.responsePrefix}", context=${JSON.stringify(opts.responsePrefixContext)}`,
+    );
+  }
   const effectivePrefix = opts.responsePrefixContext
     ? resolveResponsePrefixTemplate(opts.responsePrefix, opts.responsePrefixContext)
     : opts.responsePrefix;
+  if (opts.responsePrefix?.includes("{") && effectivePrefix !== opts.responsePrefix) {
+    logVerbose(`[responsePrefix] resolved to: "${effectivePrefix}"`);
+  }
 
   if (
     effectivePrefix &&
diff --git a/src/telegram/bot-message-dispatch.ts b/src/telegram/bot-message-dispatch.ts
index c609a07c2..22dd9a420 100644
--- a/src/telegram/bot-message-dispatch.ts
+++ b/src/telegram/bot-message-dispatch.ts
@@ -163,10 +163,14 @@ export const dispatchTelegramMessage = async ({
       disableBlockStreaming,
       onModelSelected: (ctx) => {
         // Mutate the object directly instead of reassigning to ensure the closure sees updates
+        logVerbose(
+          `[responsePrefix] telegram onModelSelected fired: provider=${ctx.provider}, model=${ctx.model}, thinkLevel=${ctx.thinkLevel}`,
+        );
         prefixContext.provider = ctx.provider;
         prefixContext.model = extractShortModelName(ctx.model);
         prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
         prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
+        logVerbose(`[responsePrefix] telegram prefixContext updated: ${JSON.stringify(prefixContext)}`);
       },
     },
   });

From e7167e35ed9f9420ff038f69cef075e06449fdff Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:29:17 -0500
Subject: [PATCH 5/6] debug: use console.log instead of logVerbose for
 always-visible logging

---
 src/auto-reply/reply/agent-runner-execution.ts | 3 ++-
 src/auto-reply/reply/normalize-reply.ts        | 9 +++++----
 src/telegram/bot-message-dispatch.ts           | 6 ++++--
 3 files changed, 11 insertions(+), 7 deletions(-)

diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index 49a50cd29..8a3033005 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -127,7 +127,8 @@ export async function runAgentTurnWithFallback(params: {
         run: (provider, model) => {
           // Notify that model selection is complete (including after fallback).
           // This allows responsePrefix template interpolation with the actual model.
-          logVerbose(
+          // eslint-disable-next-line no-console
+          console.log(
             `[responsePrefix] onModelSelected callback exists: ${!!params.opts?.onModelSelected}, provider=${provider}, model=${model}`,
           );
           params.opts?.onModelSelected?.({
diff --git a/src/auto-reply/reply/normalize-reply.ts b/src/auto-reply/reply/normalize-reply.ts
index a9d0d0060..455deb998 100644
--- a/src/auto-reply/reply/normalize-reply.ts
+++ b/src/auto-reply/reply/normalize-reply.ts
@@ -1,4 +1,3 @@
-import { logVerbose } from "../../globals.js";
 import { stripHeartbeatToken } from "../heartbeat.js";
 import { HEARTBEAT_TOKEN, isSilentReplyText, SILENT_REPLY_TOKEN } from "../tokens.js";
 import type { ReplyPayload } from "../types.js";
@@ -45,15 +44,17 @@ export function normalizeReplyPayload(
 
   // Resolve template variables in responsePrefix if context is provided
   if (opts.responsePrefix?.includes("{")) {
-    logVerbose(
+    // eslint-disable-next-line no-console
+    console.log(
       `[responsePrefix] normalizing: prefix="${opts.responsePrefix}", context=${JSON.stringify(opts.responsePrefixContext)}`,
     );
   }
   const effectivePrefix = opts.responsePrefixContext
     ? resolveResponsePrefixTemplate(opts.responsePrefix, opts.responsePrefixContext)
     : opts.responsePrefix;
-  if (opts.responsePrefix?.includes("{") && effectivePrefix !== opts.responsePrefix) {
-    logVerbose(`[responsePrefix] resolved to: "${effectivePrefix}"`);
+  if (opts.responsePrefix?.includes("{")) {
+    // eslint-disable-next-line no-console
+    console.log(`[responsePrefix] resolved to: "${effectivePrefix}"`);
   }
 
   if (
diff --git a/src/telegram/bot-message-dispatch.ts b/src/telegram/bot-message-dispatch.ts
index 22dd9a420..d005d97ae 100644
--- a/src/telegram/bot-message-dispatch.ts
+++ b/src/telegram/bot-message-dispatch.ts
@@ -163,14 +163,16 @@ export const dispatchTelegramMessage = async ({
       disableBlockStreaming,
       onModelSelected: (ctx) => {
         // Mutate the object directly instead of reassigning to ensure the closure sees updates
-        logVerbose(
+        // eslint-disable-next-line no-console
+        console.log(
           `[responsePrefix] telegram onModelSelected fired: provider=${ctx.provider}, model=${ctx.model}, thinkLevel=${ctx.thinkLevel}`,
         );
         prefixContext.provider = ctx.provider;
         prefixContext.model = extractShortModelName(ctx.model);
         prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
         prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
-        logVerbose(`[responsePrefix] telegram prefixContext updated: ${JSON.stringify(prefixContext)}`);
+        // eslint-disable-next-line no-console
+        console.log(`[responsePrefix] telegram prefixContext updated: ${JSON.stringify(prefixContext)}`);
       },
     },
   });

From 6ef3837e731ce95cbec907a400a0f585a4dedeef Mon Sep 17 00:00:00 2001
From: Sebastian <sebslight@gmail.com>
Date: Wed, 14 Jan 2026 23:36:47 -0500
Subject: [PATCH 6/6] Remove debug logging for responsePrefix template
 resolution

---
 src/auto-reply/reply/agent-runner-execution.ts |  4 ----
 src/auto-reply/reply/normalize-reply.ts        | 10 ----------
 src/telegram/bot-message-dispatch.ts           |  6 ------
 3 files changed, 20 deletions(-)

diff --git a/src/auto-reply/reply/agent-runner-execution.ts b/src/auto-reply/reply/agent-runner-execution.ts
index 8a3033005..d1f0e276d 100644
--- a/src/auto-reply/reply/agent-runner-execution.ts
+++ b/src/auto-reply/reply/agent-runner-execution.ts
@@ -127,10 +127,6 @@ export async function runAgentTurnWithFallback(params: {
         run: (provider, model) => {
           // Notify that model selection is complete (including after fallback).
           // This allows responsePrefix template interpolation with the actual model.
-          // eslint-disable-next-line no-console
-          console.log(
-            `[responsePrefix] onModelSelected callback exists: ${!!params.opts?.onModelSelected}, provider=${provider}, model=${model}`,
-          );
           params.opts?.onModelSelected?.({
             provider,
             model,
diff --git a/src/auto-reply/reply/normalize-reply.ts b/src/auto-reply/reply/normalize-reply.ts
index 455deb998..ac8e85f2e 100644
--- a/src/auto-reply/reply/normalize-reply.ts
+++ b/src/auto-reply/reply/normalize-reply.ts
@@ -43,19 +43,9 @@ export function normalizeReplyPayload(
   }
 
   // Resolve template variables in responsePrefix if context is provided
-  if (opts.responsePrefix?.includes("{")) {
-    // eslint-disable-next-line no-console
-    console.log(
-      `[responsePrefix] normalizing: prefix="${opts.responsePrefix}", context=${JSON.stringify(opts.responsePrefixContext)}`,
-    );
-  }
   const effectivePrefix = opts.responsePrefixContext
     ? resolveResponsePrefixTemplate(opts.responsePrefix, opts.responsePrefixContext)
     : opts.responsePrefix;
-  if (opts.responsePrefix?.includes("{")) {
-    // eslint-disable-next-line no-console
-    console.log(`[responsePrefix] resolved to: "${effectivePrefix}"`);
-  }
 
   if (
     effectivePrefix &&
diff --git a/src/telegram/bot-message-dispatch.ts b/src/telegram/bot-message-dispatch.ts
index d005d97ae..c609a07c2 100644
--- a/src/telegram/bot-message-dispatch.ts
+++ b/src/telegram/bot-message-dispatch.ts
@@ -163,16 +163,10 @@ export const dispatchTelegramMessage = async ({
       disableBlockStreaming,
       onModelSelected: (ctx) => {
         // Mutate the object directly instead of reassigning to ensure the closure sees updates
-        // eslint-disable-next-line no-console
-        console.log(
-          `[responsePrefix] telegram onModelSelected fired: provider=${ctx.provider}, model=${ctx.model}, thinkLevel=${ctx.thinkLevel}`,
-        );
         prefixContext.provider = ctx.provider;
         prefixContext.model = extractShortModelName(ctx.model);
         prefixContext.modelFull = `${ctx.provider}/${ctx.model}`;
         prefixContext.thinkingLevel = ctx.thinkLevel ?? "off";
-        // eslint-disable-next-line no-console
-        console.log(`[responsePrefix] telegram prefixContext updated: ${JSON.stringify(prefixContext)}`);
       },
     },
   });