@didim365/agent-cli-core 0.2.9 → 0.2.11

package/README.md

@@ -1,30 +1,31 @@
-# Gemini CLI
+# Didim Agent CLI
 
-[![Gemini CLI CI](https://github.com/google-gemini/gemini-cli/actions/workflows/ci.yml/badge.svg)](https://github.com/google-gemini/gemini-cli/actions/workflows/ci.yml)
-[![Gemini CLI E2E (Chained)](https://github.com/google-gemini/gemini-cli/actions/workflows/chained_e2e.yml/badge.svg)](https://github.com/google-gemini/gemini-cli/actions/workflows/chained_e2e.yml)
+[![CI](https://github.com/google-gemini/gemini-cli/actions/workflows/ci.yml/badge.svg)](https://github.com/google-gemini/gemini-cli/actions/workflows/ci.yml)
+[![E2E](https://github.com/google-gemini/gemini-cli/actions/workflows/chained_e2e.yml/badge.svg)](https://github.com/google-gemini/gemini-cli/actions/workflows/chained_e2e.yml)
 [![Version](https://img.shields.io/npm/v/@didim365/agent-cli)](https://www.npmjs.com/package/@didim365/agent-cli)
 [![License](https://img.shields.io/github/license/google-gemini/gemini-cli)](https://github.com/google-gemini/gemini-cli/blob/main/LICENSE)
-[![View Code Wiki](https://assets.codewiki.google/readme-badge/static.svg)](https://codewiki.google/github.com/google-gemini/gemini-cli?utm_source=badge&utm_medium=github&utm_campaign=github.com/google-gemini/gemini-cli)
 
-![Gemini CLI Screenshot](./docs/assets/gemini-screenshot.png)
+![Didim Agent CLI Screenshot](./docs/assets/gemini-screenshot.png)
 
-Gemini CLI is an open-source AI agent that brings the power of multiple AI
-providers directly into your terminal. It supports **Gemini**, **Claude**,
-**OpenAI**, and **OpenAI-compatible** (vLLM, Ollama, LM Studio) endpoints,
-giving you the most direct path from your prompt to your preferred model.
+Didim Agent CLI is an open-source AI agent that brings the power of multiple AI
+providers directly into your terminal. Built on the Gemini CLI foundation, it
+supports **Gemini**, **Claude**, **OpenAI**, and **OpenAI-compatible** (vLLM,
+Ollama, LM Studio) endpoints through a unified **provider adapter
+architecture**, giving you the most direct path from your prompt to your
+preferred model.
 
-Learn all about Gemini CLI in our [documentation](https://geminicli.com/docs/).
+Learn all about Didim Agent CLI in our [documentation](./docs/index.md).
 
-## 🚀 Why Gemini CLI?
+## 🚀 Why Didim Agent CLI?
 
-- **🎯 Free tier**: 60 requests/min and 1,000 requests/day with personal Google
-  account.
 - **🧠 Multi-provider support**: Use Gemini, Claude, OpenAI, or local models
-  (vLLM/Ollama) — switch providers and models with `/model`.
+  (vLLM/Ollama) — switch providers and models with `/model` or `/auth login`.
 - **🔧 Built-in tools**: Google Search grounding, file operations, shell
-  commands, web fetching.
-- **🔌 Extensible**: MCP (Model Context Protocol) support for custom
-  integrations.
+  commands, web fetching — all tools work across providers.
+- **🔌 Extensible**: MCP (Model Context Protocol) support with deterministic
+  tool naming and sLM-compatible parameter normalization.
+- **🤖 Sub-agent support**: Sub-agents work with all providers via the
+  provider-independent `llm*` pipeline.
 - **💻 Terminal-first**: Designed for developers who live in the command line.
 - **🛡️ Open source**: Apache 2.0 licensed.
 
@@ -129,7 +130,7 @@ npm install -g @didim365/agent-cli@nightly
   [Google Search](https://ai.google.dev/gemini-api/docs/grounding) for real-time
   information
 - Conversation checkpointing to save and resume complex sessions
-- Custom context files (GEMINI.md) to tailor behavior for your projects
+- Custom context files (AGENTS.md) to tailor behavior for your projects
 
 ### GitHub Integration
 
@@ -151,12 +152,16 @@ Choose the authentication method that best fits your needs. You can also use
 `/auth login` inside the CLI to interactively select a provider and enter your
 API key.
 
+> **Note:** Both `DIDIM_*` and `GEMINI_*` environment variable prefixes are
+> supported. The CLI uses a central `resolveEnv()` utility that checks `DIDIM_*`
+> first, then falls back to `GEMINI_*` for backward compatibility.
+
 ### Option 1: Login with Google (Gemini)
 
 **✨ Best for:** Individual developers and Gemini Code Assist license holders.
 
 ```bash
-gemini
+didim
 # Select "Login with Google" and follow the browser authentication flow
 ```
 
@@ -164,7 +169,7 @@ For organization accounts, set your Google Cloud project first:
 
 ```bash
 export GOOGLE_CLOUD_PROJECT="YOUR_PROJECT_ID"
-gemini
+didim
 ```
 
 ### Option 2: Gemini API Key
@@ -173,7 +178,7 @@ gemini
 
 ```bash
 export GEMINI_API_KEY="YOUR_API_KEY"
-gemini
+didim
 ```
 
 ### Option 3: Claude (Anthropic)
@@ -182,7 +187,7 @@ gemini
 
 ```bash
 export ANTHROPIC_API_KEY="YOUR_API_KEY"
-gemini
+didim
 ```
 
 ### Option 4: OpenAI
@@ -191,7 +196,7 @@ gemini
 
 ```bash
 export OPENAI_API_KEY="YOUR_API_KEY"
-gemini
+didim
 ```
 
 ### Option 5: Vertex AI
@@ -201,7 +206,7 @@ gemini
 ```bash
 export GOOGLE_API_KEY="YOUR_API_KEY"
 export GOOGLE_GENAI_USE_VERTEXAI=true
-gemini
+didim
 ```
 
 ### Option 6: OpenAI-compatible (vLLM, Ollama, LM Studio)
@@ -212,7 +217,7 @@ gemini
 export ENABLE_MULTI_PROVIDER=true
 export LLM_PROVIDER=openai-compatible
 export LLM_BASE_URL="http://localhost:8000/v1"
-gemini
+didim
 ```
 
 For detailed setup for each provider, see the
@@ -226,21 +231,21 @@ For detailed setup for each provider, see the
 #### Start in current directory
 
 ```bash
-gemini
+didim
 ```
 
 #### Include multiple directories
 
 ```bash
-gemini --include-directories ../lib,../docs
+didim --include-directories ../lib,../docs
 ```
 
 #### Use specific model
 
 ```bash
-gemini -m gemini-2.5-flash          # Gemini
-gemini -m claude-sonnet-4-5-20250929  # Claude
-gemini -m gpt-4.1                    # OpenAI
+didim -m gemini-2.5-flash            # Gemini
+didim -m claude-sonnet-4-5-20250929  # Claude
+didim -m gpt-4.1                     # OpenAI
 ```
 
 #### Non-interactive mode for scripts
@@ -248,21 +253,21 @@ gemini -m gpt-4.1                    # OpenAI
 Get a simple text response:
 
 ```bash
-gemini -p "Explain the architecture of this codebase"
+didim -p "Explain the architecture of this codebase"
 ```
 
 For more advanced scripting, including how to parse JSON and handle errors, use
 the `--output-format json` flag to get structured output:
 
 ```bash
-gemini -p "Explain the architecture of this codebase" --output-format json
+didim -p "Explain the architecture of this codebase" --output-format json
 ```
 
 For real-time event streaming (useful for monitoring long-running operations),
 use `--output-format stream-json` to get newline-delimited JSON events:
 
 ```bash
-gemini -p "Run tests and deploy" --output-format stream-json
+didim -p "Run tests and deploy" --output-format stream-json
 ```
 
 ### Quick Examples
@@ -271,16 +276,16 @@ gemini -p "Run tests and deploy" --output-format stream-json
 
 ```bash
 cd new-project/
-gemini
+didim
 > Write me a Discord bot that answers questions using a FAQ.md file I will provide
 ```
 
 #### Analyze existing code
 
 ```bash
-git clone https://github.com/google-gemini/gemini-cli
-cd gemini-cli
-gemini
+git clone https://github.com/user/project
+cd project
+didim
 > Give me a summary of all of the changes that went in yesterday
 ```
 
@@ -303,8 +308,8 @@ gemini
   (`/help`, `/chat`, etc).
 - [**Custom Commands**](./docs/cli/custom-commands.md) - Create your own
   reusable commands.
-- [**Context Files (GEMINI.md)**](./docs/cli/gemini-md.md) - Provide persistent
-  context to Gemini CLI.
+- [**Context Files (AGENTS.md)**](./docs/cli/gemini-md.md) - Provide persistent
+  context to the CLI.
 - [**Checkpointing**](./docs/cli/checkpointing.md) - Save and resume
   conversations.
 - [**Token Caching**](./docs/cli/token-caching.md) - Optimize token usage.
@@ -352,7 +357,7 @@ export ENABLE_MULTI_PROVIDER=true
 export LLM_PROVIDER=openai-compatible
 export LLM_BASE_URL="http://localhost:8000/v1"
 export LLM_MODEL="Qwen/Qwen2.5-7B-Instruct"
-gemini -m Qwen/Qwen2.5-7B-Instruct
+didim -m Qwen/Qwen2.5-7B-Instruct
 ```
 
 ### Troubleshooting & Support
@@ -364,8 +369,8 @@ gemini -m Qwen/Qwen2.5-7B-Instruct
 
 ### Using MCP Servers
 
-Configure MCP servers in `~/.gemini/settings.json` to extend Gemini CLI with
-custom tools:
+Configure MCP servers in `~/.didim/settings.json` (or `~/.gemini/settings.json`
+for backward compatibility) to extend the CLI with custom tools:
 
 ```text
 > @github List my open pull requests
@@ -373,6 +378,11 @@ custom tools:
 > @database Run a query to find inactive users
 ```
 
+MCP tool naming is **deterministic** — tools are registered with consistent
+names regardless of server discovery order. Tool parameters are automatically
+normalized via schema-based coercion, with enhanced tolerance for sLM (small
+Language Model) tool call formatting.
+
 See the [MCP Server Integration guide](./docs/tools/mcp-server.md) for setup
 instructions.
 
@@ -416,5 +426,5 @@ See the [Uninstall Guide](docs/cli/uninstall.md) for removal instructions.
 ---
 
 <p align="center">
-  Built with ❤️ by Google and the open source community
+  Built on Gemini CLI by Google — extended by Didim365
 </p>

package/dist/src/config/costEstimation.d.ts

@@ -0,0 +1,65 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+/**
+ * Pricing information for a single model.
+ *
+ * All prices are in USD per 1 million tokens.
+ */
+export interface ModelPricing {
+    /** USD per 1M input tokens */
+    inputPerMToken: number;
+    /** USD per 1M output tokens */
+    outputPerMToken: number;
+    /** USD per 1M cached input tokens (optional — fallback to inputPerMToken) */
+    cachedPerMToken?: number;
+}
+/**
+ * Static pricing table: provider → model → pricing.
+ *
+ * Prices sourced from official API pricing pages (as of 2026-02).
+ * Models not listed here will be treated as $0 (unknown pricing).
+ */
+export declare const MODEL_PRICING: Record<string, Record<string, ModelPricing>>;
+/**
+ * Result of cost estimation across all models.
+ */
+export interface CostEstimate {
+    /** Total estimated cost in USD */
+    totalCost: number;
+    /** Cost breakdown per provider in USD */
+    byProvider: Record<string, number>;
+}
+/**
+ * Token shape expected by estimateCost — subset of ModelMetrics.tokens.
+ */
+interface TokenInfo {
+    input: number;
+    cached: number;
+    candidates: number;
+}
+/**
+ * Estimate cost based on token usage and static pricing table.
+ *
+ * @param models - Record of composite-key → token info
+ * @param parseKey - Function to extract { provider, model } from composite key
+ * @returns CostEstimate with total and per-provider breakdown
+ */
+export declare function estimateCost(models: Record<string, {
+    tokens: TokenInfo;
+}>, parseKey: (key: string) => {
+    provider: string;
+    model: string;
+}): CostEstimate;
+/**
+ * Format a CostEstimate into a human-readable string.
+ *
+ * - Zero: "$0.00"
+ * - Very small (< $0.01): "< $0.01"
+ * - Single provider: "$1.23"
+ * - Multi-provider: "$5.68 (Gemini $2.35 + Claude $3.33)"
+ */
+export declare function formatCostString(estimate: CostEstimate): string;
+export {};

package/dist/src/config/costEstimation.js

@@ -0,0 +1,133 @@
+/**
+ * @license
+ * Copyright 2026 Google LLC
+ * SPDX-License-Identifier: Apache-2.0
+ */
+/**
+ * Static pricing table: provider → model → pricing.
+ *
+ * Prices sourced from official API pricing pages (as of 2026-02).
+ * Models not listed here will be treated as $0 (unknown pricing).
+ */
+export const MODEL_PRICING = {
+    gemini: {
+        'gemini-2.5-pro': {
+            inputPerMToken: 1.25,
+            outputPerMToken: 10.0,
+            cachedPerMToken: 0.125,
+        },
+        'gemini-2.5-flash': {
+            inputPerMToken: 0.3,
+            outputPerMToken: 2.5,
+            cachedPerMToken: 0.03,
+        },
+        'gemini-2.5-flash-lite': {
+            inputPerMToken: 0.1,
+            outputPerMToken: 0.4,
+            cachedPerMToken: 0.01,
+        },
+    },
+    claude: {
+        'claude-opus-4-6': {
+            inputPerMToken: 5.0,
+            outputPerMToken: 25.0,
+            cachedPerMToken: 0.5,
+        },
+        'claude-sonnet-4-5-20250929': {
+            inputPerMToken: 3.0,
+            outputPerMToken: 15.0,
+            cachedPerMToken: 0.3,
+        },
+        'claude-haiku-4-5-20251001': {
+            inputPerMToken: 1.0,
+            outputPerMToken: 5.0,
+            cachedPerMToken: 0.1,
+        },
+    },
+    openai: {
+        'gpt-5.2': {
+            inputPerMToken: 1.25,
+            outputPerMToken: 10.0,
+            cachedPerMToken: 0.625,
+        },
+        'gpt-5-mini': {
+            inputPerMToken: 0.4,
+            outputPerMToken: 1.6,
+            cachedPerMToken: 0.1,
+        },
+        'gpt-4.1': {
+            inputPerMToken: 2.0,
+            outputPerMToken: 8.0,
+            cachedPerMToken: 0.5,
+        },
+        'gpt-4.1-mini': {
+            inputPerMToken: 0.4,
+            outputPerMToken: 1.6,
+            cachedPerMToken: 0.1,
+        },
+        o3: {
+            inputPerMToken: 2.0,
+            outputPerMToken: 8.0,
+        },
+        'o4-mini': {
+            inputPerMToken: 1.1,
+            outputPerMToken: 4.4,
+            cachedPerMToken: 0.275,
+        },
+    },
+};
+const TOKENS_PER_MILLION = 1_000_000;
+/**
+ * Estimate cost based on token usage and static pricing table.
+ *
+ * @param models - Record of composite-key → token info
+ * @param parseKey - Function to extract { provider, model } from composite key
+ * @returns CostEstimate with total and per-provider breakdown
+ */
+export function estimateCost(models, parseKey) {
+    const byProvider = {};
+    for (const [key, entry] of Object.entries(models)) {
+        const { provider, model } = parseKey(key);
+        const pricing = MODEL_PRICING[provider]?.[model];
+        if (!pricing)
+            continue;
+        const nonCachedInput = Math.max(0, entry.tokens.input - entry.tokens.cached);
+        const cachedRate = pricing.cachedPerMToken ?? pricing.inputPerMToken;
+        const inputCost = (nonCachedInput / TOKENS_PER_MILLION) * pricing.inputPerMToken;
+        const cachedCost = (entry.tokens.cached / TOKENS_PER_MILLION) * cachedRate;
+        const outputCost = (entry.tokens.candidates / TOKENS_PER_MILLION) * pricing.outputPerMToken;
+        const modelCost = inputCost + cachedCost + outputCost;
+        byProvider[provider] = (byProvider[provider] ?? 0) + modelCost;
+    }
+    const totalCost = Object.values(byProvider).reduce((sum, v) => sum + v, 0);
+    return { totalCost, byProvider };
+}
+/**
+ * Capitalize first letter of a string.
+ */
+function capitalize(s) {
+    return s.charAt(0).toUpperCase() + s.slice(1);
+}
+/**
+ * Format a CostEstimate into a human-readable string.
+ *
+ * - Zero: "$0.00"
+ * - Very small (< $0.01): "< $0.01"
+ * - Single provider: "$1.23"
+ * - Multi-provider: "$5.68 (Gemini $2.35 + Claude $3.33)"
+ */
+export function formatCostString(estimate) {
+    if (estimate.totalCost === 0)
+        return '$0.00';
+    if (estimate.totalCost > 0 && estimate.totalCost < 0.01)
+        return '< $0.01';
+    const providers = Object.entries(estimate.byProvider);
+    const totalStr = `$${estimate.totalCost.toFixed(2)}`;
+    if (providers.length <= 1)
+        return totalStr;
+    const breakdown = providers
+        .map(([name, cost]) => `${capitalize(name)} $${cost.toFixed(2)}`)
+        .join(' + ');
+    return `${totalStr} (${breakdown})`;
+}
+//# sourceMappingURL=costEstimation.js.map

package/dist/src/config/costEstimation.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"costEstimation.js","sourceRoot":"","sources":["../../../src/config/costEstimation.ts"],"names":[],"mappings":"AAAA;;;;GAIG;AAgBH;;;;;GAKG;AACH,MAAM,CAAC,MAAM,aAAa,GAAiD;IACzE,MAAM,EAAE;QACN,gBAAgB,EAAE;YAChB,cAAc,EAAE,IAAI;YACpB,eAAe,EAAE,IAAI;YACrB,eAAe,EAAE,KAAK;SACvB;QACD,kBAAkB,EAAE;YAClB,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,IAAI;SACtB;QACD,uBAAuB,EAAE;YACvB,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,IAAI;SACtB;KACF;IACD,MAAM,EAAE;QACN,iBAAiB,EAAE;YACjB,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,IAAI;YACrB,eAAe,EAAE,GAAG;SACrB;QACD,4BAA4B,EAAE;YAC5B,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,IAAI;YACrB,eAAe,EAAE,GAAG;SACrB;QACD,2BAA2B,EAAE;YAC3B,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,GAAG;SACrB;KACF;IACD,MAAM,EAAE;QACN,SAAS,EAAE;YACT,cAAc,EAAE,IAAI;YACpB,eAAe,EAAE,IAAI;YACrB,eAAe,EAAE,KAAK;SACvB;QACD,YAAY,EAAE;YACZ,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,GAAG;SACrB;QACD,SAAS,EAAE;YACT,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,GAAG;SACrB;QACD,cAAc,EAAE;YACd,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,GAAG;SACrB;QACD,EAAE,EAAE;YACF,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;SACrB;QACD,SAAS,EAAE;YACT,cAAc,EAAE,GAAG;YACnB,eAAe,EAAE,GAAG;YACpB,eAAe,EAAE,KAAK;SACvB;KACF;CACF,CAAC;AAqBF,MAAM,kBAAkB,GAAG,SAAS,CAAC;AAErC;;;;;;GAMG;AACH,MAAM,UAAU,YAAY,CAC1B,MAA6C,EAC7C,QAA8D;IAE9D,MAAM,UAAU,GAA2B,EAAE,CAAC;IAE9C,KAAK,MAAM,CAAC,GAAG,EAAE,KAAK,CAAC,IAAI,MAAM,CAAC,OAAO,CAAC,MAAM,CAAC,EAAE,CAAC;QAClD,MAAM,EAAE,QAAQ,EAAE,KAAK,EAAE,GAAG,QAAQ,CAAC,GAAG,CAAC,CAAC;QAC1C,MAAM,OAAO,GAAG,aAAa,CAAC,QAAQ,CAAC,EAAE,CAAC,KAAK,CAAC,CAAC;QACjD,IAAI,CAAC,OAAO;YAAE,SAAS;QAEvB,MAAM,cAAc,GAAG,IAAI,CAAC,GAAG,CAAC,CAAC,EAAE,KAAK,CAAC,MAAM,CAAC,KAAK,GAAG,KAAK,CAAC,MAAM,CAAC,MAAM,CAAC,CAAC;QAC7E,MAAM,UAAU,GAAG,OAAO,CAAC,eAAe,IAAI,OAAO,CAAC,cAAc,CAAC;QAErE,MAAM,SAAS,GACb,CAAC,cAAc,GAAG,kBAAkB,CAAC,GAAG,OAAO,CAAC,cAAc,CAAC;QACjE,MAAM,UAAU,GACd,CAAC,KAAK,CAAC,MAAM,CAAC,MAAM,GAAG,kBAAkB,CAAC,GAAG,UAAU,CAAC;QAC1D,MAAM,UAAU,GACd,CAAC,KAAK,CAAC,MAAM,CAAC,UAAU,GAAG,kBAAkB,CAAC,GAAG,OAAO,CAAC,eAAe,CAAC;QAE3E,MAAM,SAAS,GAAG,SAAS,GAAG,UAAU,GAAG,UAAU,CAAC;QACtD,UAAU,CAAC,QAAQ,CAAC,GAAG,CAAC,UAAU,CAAC,QAAQ,CAAC,IAAI,CAAC,CAAC,GAAG,SAAS,CAAC;IACjE,CAAC;IAED,MAAM,SAAS,GAAG,MAAM,CAAC,MAAM,CAAC,UAAU,CAAC,CAAC,MAAM,CAAC,CAAC,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC,GAAG,GAAG,CAAC,EAAE,CAAC,CAAC,CAAC;IAC3E,OAAO,EAAE,SAAS,EAAE,UAAU,EAAE,CAAC;AACnC,CAAC;AAED;;GAEG;AACH,SAAS,UAAU,CAAC,CAAS;IAC3B,OAAO,CAAC,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC,WAAW,EAAE,GAAG,CAAC,CAAC,KAAK,CAAC,CAAC,CAAC,CAAC;AAChD,CAAC;AAED;;;;;;;GAOG;AACH,MAAM,UAAU,gBAAgB,CAAC,QAAsB;IACrD,IAAI,QAAQ,CAAC,SAAS,KAAK,CAAC;QAAE,OAAO,OAAO,CAAC;IAC7C,IAAI,QAAQ,CAAC,SAAS,GAAG,CAAC,IAAI,QAAQ,CAAC,SAAS,GAAG,IAAI;QAAE,OAAO,SAAS,CAAC;IAE1E,MAAM,SAAS,GAAG,MAAM,CAAC,OAAO,CAAC,QAAQ,CAAC,UAAU,CAAC,CAAC;IACtD,MAAM,QAAQ,GAAG,IAAI,QAAQ,CAAC,SAAS,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC;IAErD,IAAI,SAAS,CAAC,MAAM,IAAI,CAAC;QAAE,OAAO,QAAQ,CAAC;IAE3C,MAAM,SAAS,GAAG,SAAS;SACxB,GAAG,CAAC,CAAC,CAAC,IAAI,EAAE,IAAI,CAAC,EAAE,EAAE,CAAC,GAAG,UAAU,CAAC,IAAI,CAAC,KAAK,IAAI,CAAC,OAAO,CAAC,CAAC,CAAC,EAAE,CAAC;SAChE,IAAI,CAAC,KAAK,CAAC,CAAC;IAEf,OAAO,GAAG,QAAQ,KAAK,SAAS,GAAG,CAAC;AACtC,CAAC"}

package/dist/src/core/loggingContentGenerator.d.ts

@@ -8,13 +8,17 @@ import type { Config } from '../config/config.js';
 import type { UserTierId } from '../code_assist/types.js';
 import type { LlmGenerateRequest, LlmGenerateResponse, LlmTokenCount, GenerateOptions } from '../providers/types.js';
 import type { LlmEventStream } from '../providers/events.js';
+import type { ProviderQuotaService } from '../telemetry/providerQuotaService.js';
 /**
  * A decorator that wraps a ContentGenerator to add logging to API calls.
  */
 export declare class LoggingContentGenerator implements ContentGenerator {
     private readonly wrapped;
     private readonly config;
-    constructor(wrapped: ContentGenerator, config: Config);
+    private providerQuotaService?;
+    constructor(wrapped: ContentGenerator, config: Config, providerQuotaService?: ProviderQuotaService);
+    /** Set the ProviderQuotaService after construction (for late binding). */
+    setProviderQuotaService(service: ProviderQuotaService): void;
     getWrapped(): ContentGenerator;
     get userTier(): UserTierId | undefined;
     get userTierName(): string | undefined;
@@ -31,5 +35,7 @@ export declare class LoggingContentGenerator implements ContentGenerator {
     llmGenerateContent(request: LlmGenerateRequest, userPromptId: string, options?: GenerateOptions): Promise<LlmGenerateResponse>;
     llmGenerateContentStream(request: LlmGenerateRequest, userPromptId: string, options?: GenerateOptions): LlmEventStream;
     private llmLoggingStreamWrapper;
+    private _logLlmApiResponse;
+    private _logLlmApiError;
     llmCountTokens(request: LlmGenerateRequest): Promise<LlmTokenCount>;
 }

package/dist/src/core/loggingContentGenerator.js

@@ -4,7 +4,9 @@
  * SPDX-License-Identifier: Apache-2.0
  */
 import { ApiRequestEvent, ApiResponseEvent, ApiErrorEvent, } from '../telemetry/types.js';
-import { logApiError, logApiRequest, logApiResponse, } from '../telemetry/loggers.js';
+import { logApiError, logApiRequest, logApiResponse, logProviderApiResponse, logProviderApiError, } from '../telemetry/loggers.js';
+import { LlmEventType } from '../providers/events.js';
+import { createProviderApiResponseEvent, createProviderApiErrorEvent, } from '../providers/telemetryBridge.js';
 import { debugLogger } from '../utils/debugLogger.js';
 import { CodeAssistServer } from '../code_assist/server.js';
 import { toContents } from '../code_assist/converter.js';
@@ -16,9 +18,15 @@ import { runInDevTraceSpan } from '../telemetry/trace.js';
 export class LoggingContentGenerator {
     wrapped;
     config;
-    constructor(wrapped, config) {
+    providerQuotaService;
+    constructor(wrapped, config, providerQuotaService) {
         this.wrapped = wrapped;
         this.config = config;
+        this.providerQuotaService = providerQuotaService;
+    }
+    /** Set the ProviderQuotaService after construction (for late binding). */
+    setProviderQuotaService(service) {
+        this.providerQuotaService = service;
     }
     getWrapped() {
         return this.wrapped;
@@ -208,16 +216,23 @@ export class LoggingContentGenerator {
             throw new Error('Wrapped generator does not support provider-independent API');
         }
         const startTime = Date.now();
+        const provider = this.wrapped.providerName ?? 'unknown';
         debugLogger.debug(`[LLM] generateContent model=${request.model} promptId=${userPromptId}`);
         try {
             const response = await this.wrapped.llmGenerateContent(request, userPromptId, options);
             const durationMs = Date.now() - startTime;
             debugLogger.debug(`[LLM] generateContent completed in ${durationMs}ms`);
+            this._logLlmApiResponse(request.model, durationMs, userPromptId, provider, response.usage);
+            // Bridge rate-limit data to ProviderQuotaService (non-stream path)
+            if (response.rateLimits && this.providerQuotaService) {
+                this.providerQuotaService.update(provider, response.rateLimits);
+            }
             return response;
         }
         catch (error) {
             const durationMs = Date.now() - startTime;
             debugLogger.debug(`[LLM] generateContent error after ${durationMs}ms: ${error}`);
+            this._logLlmApiError(request.model, durationMs, userPromptId, provider, error);
             throw error;
         }
     }
@@ -226,24 +241,85 @@ export class LoggingContentGenerator {
             throw new Error('Wrapped generator does not support provider-independent API');
         }
         debugLogger.debug(`[LLM] generateContentStream model=${request.model} promptId=${userPromptId}`);
+        const startTime = Date.now();
         const stream = this.wrapped.llmGenerateContentStream(request, userPromptId, options);
-        return this.llmLoggingStreamWrapper(stream, request.model, userPromptId);
+        return this.llmLoggingStreamWrapper(stream, request.model, userPromptId, startTime);
     }
-    async *llmLoggingStreamWrapper(stream, model, userPromptId) {
-        const startTime = Date.now();
+    async *llmLoggingStreamWrapper(stream, model, userPromptId, startTime) {
+        const provider = this.wrapped.providerName ?? 'unknown';
+        let collectedUsage;
+        let hasError = false;
         try {
             for await (const event of stream) {
+                // Collect usage from MessageEnd or Finished event.
+                // OpenAI puts usage in MessageEnd; Claude puts usage in Finished.
+                if ((event.type === LlmEventType.MessageEnd ||
+                    event.type === LlmEventType.Finished) &&
+                    event.usage) {
+                    collectedUsage = event.usage;
+                }
+                // Bridge rate-limit data to ProviderQuotaService
+                if (event.type === LlmEventType.MessageEnd) {
+                    const rateLimits = event
+                        .rateLimits;
+                    if (rateLimits && this.providerQuotaService) {
+                        this.providerQuotaService.update(provider, rateLimits);
+                    }
+                }
+                // Detect Error events (yielded, not thrown)
+                if (event.type === LlmEventType.Error) {
+                    hasError = true;
+                    const durationMs = Date.now() - startTime;
+                    this._logLlmApiError(model, durationMs, userPromptId, provider, event.error);
+                }
                 yield event;
             }
             const durationMs = Date.now() - startTime;
             debugLogger.debug(`[LLM] generateContentStream completed in ${durationMs}ms`);
+            // Log response telemetry only if no error event was encountered
+            if (!hasError) {
+                this._logLlmApiResponse(model, durationMs, userPromptId, provider, collectedUsage);
+            }
         }
         catch (error) {
             const durationMs = Date.now() - startTime;
             debugLogger.debug(`[LLM] generateContentStream error after ${durationMs}ms model=${model} promptId=${userPromptId}: ${error}`);
+            // Skip error telemetry if:
+            // - Error was already logged via yielded Error event (prevents double recording)
+            // - User cancelled the operation (AbortError is not an API failure)
+            const isAbort = error instanceof Error && error.name === 'AbortError';
+            if (!hasError && !isAbort) {
+                this._logLlmApiError(model, durationMs, userPromptId, provider, error);
+            }
             throw error;
         }
     }
+    _logLlmApiResponse(model, durationMs, promptId, provider, usage) {
+        const defaultUsage = {
+            promptTokens: 0,
+            completionTokens: 0,
+            totalTokens: 0,
+        };
+        const event = createProviderApiResponseEvent({
+            model,
+            durationMs,
+            promptId,
+            usage: usage ?? defaultUsage,
+            provider,
+        });
+        logProviderApiResponse(this.config, event);
+    }
+    _logLlmApiError(model, durationMs, promptId, provider, error) {
+        const errorMsg = error instanceof Error ? error.message : String(error ?? 'Unknown error');
+        const event = createProviderApiErrorEvent({
+            model,
+            error: errorMsg,
+            durationMs,
+            promptId,
+            provider,
+        });
+        logProviderApiError(this.config, event);
+    }
     async llmCountTokens(request) {
         if (!this.wrapped.llmCountTokens) {
             throw new Error('Wrapped generator does not support provider-independent API');