Files
automaker/apps/server/src/providers/simple-query-service.ts
Stefan de Vogelaere a1f234c7e2 feat: Claude Compatible Providers System (#629)
* feat: refactor Claude API Profiles to Claude Compatible Providers

- Rename ClaudeApiProfile to ClaudeCompatibleProvider with models[] array
- Each ProviderModel has mapsToClaudeModel field for Claude tier mapping
- Add providerType field for provider-specific icons (glm, minimax, openrouter)
- Add thinking level support for provider models in phase selectors
- Show all mapped Claude models per provider model (e.g., "Maps to Haiku, Sonnet, Opus")
- Add Bulk Replace feature to switch all phases to a provider at once
- Hide Bulk Replace button when no providers are enabled
- Fix project-level phaseModelOverrides not persisting after refresh
- Fix deleting last provider not persisting (remove empty array guard)
- Add getProviderByModelId() helper for all SDK routes
- Update all routes to pass provider config for provider models
- Update terminology from "profiles" to "providers" throughout UI
- Update documentation to reflect new provider system

* fix: atomic writer race condition and bulk replace reset to defaults

1. AtomicWriter Race Condition Fix (libs/utils/src/atomic-writer.ts):
   - Changed temp file naming from Date.now() to Date.now() + random hex
   - Uses crypto.randomBytes(4).toString('hex') for uniqueness
   - Prevents ENOENT errors when multiple concurrent writes happen
     within the same millisecond

2. Bulk Replace "Anthropic Direct" Reset (both dialogs):
   - When selecting "Anthropic Direct", now uses DEFAULT_PHASE_MODELS
   - Properly resets thinking levels and other settings to defaults
   - Added thinkingLevel to the change detection comparison
   - Affects both global and project-level bulk replace dialogs

* fix: update tests for new model resolver passthrough behavior

1. model-resolver tests:
   - Unknown models now pass through unchanged (provider model support)
   - Removed expectations for warnings on unknown models
   - Updated case sensitivity and edge case tests accordingly
   - Added tests for provider-like model names (GLM-4.7, MiniMax-M2.1)

2. atomic-writer tests:
   - Updated regex to match new temp file format with random suffix
   - Format changed from .tmp.{timestamp} to .tmp.{timestamp}.{hex}

* refactor: simplify getPhaseModelWithOverrides calls per code review

Address code review feedback on PR #629:
- Make settingsService parameter optional in getPhaseModelWithOverrides
- Function now handles undefined settingsService gracefully by returning defaults
- Remove redundant ternary checks in 4 call sites:
  - apps/server/src/routes/context/routes/describe-file.ts
  - apps/server/src/routes/context/routes/describe-image.ts
  - apps/server/src/routes/worktree/routes/generate-commit-message.ts
  - apps/server/src/services/auto-mode-service.ts
- Remove unused DEFAULT_PHASE_MODELS imports where applicable

* test: fix server tests for provider model passthrough behavior

- Update model-resolver.test.ts to expect unknown models to pass through
  unchanged (supports ClaudeCompatibleProvider models like GLM-4.7)
- Remove warning expectations for unknown models (valid for providers)
- Add missing getCredentials and getGlobalSettings mocks to
  ideation-service.test.ts for settingsService

* fix: address code review feedback for model providers

- Honor thinkingLevel in generate-commit-message.ts
- Pass claudeCompatibleProvider in ideation-service.ts for provider models
- Resolve provider configuration for model overrides in generate-suggestions.ts
- Update "Active Profile" to "Active Provider" label in project-claude-section
- Use substring instead of deprecated substr in api-profiles-section
- Preserve provider enabled state when editing in api-profiles-section

* fix: address CodeRabbit review issues for Claude Compatible Providers

- Fix TypeScript TS2339 error in generate-suggestions.ts where
  settingsService was narrowed to 'never' type in else branch
- Use DEFAULT_PHASE_MODELS per-phase defaults instead of hardcoded
  'sonnet' in settings-helpers.ts
- Remove duplicate eventHooks key in use-settings-migration.ts
- Add claudeCompatibleProviders to localStorage migration parsing
  and merging functions
- Handle canonical claude-* model IDs (claude-haiku, claude-sonnet,
  claude-opus) in project-models-section display names

This resolves the CI build failures and addresses code review feedback.

* fix: skip broken list-view-priority E2E test and add Priority column label

- Skip list-view-priority.spec.ts with TODO explaining the infrastructure
  issue: setupRealProject only sets localStorage but server settings
  take precedence with localStorageMigrated: true
- Add 'Priority' label to list-header.tsx for the priority column
  (was empty string, now shows proper header text)
- Increase column width to accommodate the label

The E2E test issue is that tests create features in a temp directory,
but the server loads from the E2E Test Project fixture path set in
setup-e2e-fixtures.mjs. Needs infrastructure fix to properly switch
projects or create features through UI instead of on disk.
2026-01-20 20:57:23 +01:00

276 lines
9.5 KiB
TypeScript

/**
* Simple Query Service - Simplified interface for basic AI queries
*
* Use this for routes that need simple text responses without
* complex event handling. This service abstracts away the provider
* selection and streaming details, providing a clean interface
* for common query patterns.
*
* Benefits:
* - No direct SDK imports needed in route files
* - Consistent provider routing based on model
* - Automatic text extraction from streaming responses
* - Structured output support for JSON schema responses
* - Eliminates duplicate extractTextFromStream() functions
*/
import { ProviderFactory } from './provider-factory.js';
import type {
ProviderMessage,
ContentBlock,
ThinkingLevel,
ReasoningEffort,
ClaudeApiProfile,
ClaudeCompatibleProvider,
Credentials,
} from '@automaker/types';
import { stripProviderPrefix } from '@automaker/types';
/**
* Options for simple query execution
*/
export interface SimpleQueryOptions {
/** The prompt to send to the AI (can be text or multi-part content) */
prompt: string | Array<{ type: string; text?: string; source?: object }>;
/** Model to use (with or without provider prefix) */
model?: string;
/** Working directory for the query */
cwd: string;
/** System prompt (combined with user prompt for some providers) */
systemPrompt?: string;
/** Maximum turns for agentic operations (default: 1) */
maxTurns?: number;
/** Tools to allow (default: [] for simple queries) */
allowedTools?: string[];
/** Abort controller for cancellation */
abortController?: AbortController;
/** Structured output format for JSON responses */
outputFormat?: {
type: 'json_schema';
schema: Record<string, unknown>;
};
/** Thinking level for Claude models */
thinkingLevel?: ThinkingLevel;
/** Reasoning effort for Codex/OpenAI models */
reasoningEffort?: ReasoningEffort;
/** If true, runs in read-only mode (no file writes) */
readOnly?: boolean;
/** Setting sources for CLAUDE.md loading */
settingSources?: Array<'user' | 'project' | 'local'>;
/**
* Active Claude API profile for alternative endpoint configuration
* @deprecated Use claudeCompatibleProvider instead
*/
claudeApiProfile?: ClaudeApiProfile;
/**
* Claude-compatible provider for alternative endpoint configuration.
* Takes precedence over claudeApiProfile if both are set.
*/
claudeCompatibleProvider?: ClaudeCompatibleProvider;
/** Credentials for resolving 'credentials' apiKeySource in Claude API profiles/providers */
credentials?: Credentials;
}
/**
* Result from a simple query
*/
export interface SimpleQueryResult {
/** The accumulated text response */
text: string;
/** Structured output if outputFormat was specified and provider supports it */
structured_output?: Record<string, unknown>;
}
/**
* Options for streaming query execution
*/
export interface StreamingQueryOptions extends SimpleQueryOptions {
/** Callback for each text chunk received */
onText?: (text: string) => void;
/** Callback for tool use events */
onToolUse?: (tool: string, input: unknown) => void;
/** Callback for thinking blocks (if available) */
onThinking?: (thinking: string) => void;
}
/**
* Default model to use when none specified
*/
const DEFAULT_MODEL = 'claude-sonnet-4-20250514';
/**
* Execute a simple query and return the text result
*
* Use this for simple, non-streaming queries where you just need
* the final text response. For more complex use cases with progress
* callbacks, use streamingQuery() instead.
*
* @example
* ```typescript
* const result = await simpleQuery({
* prompt: 'Generate a title for: user authentication',
* cwd: process.cwd(),
* systemPrompt: 'You are a title generator...',
* maxTurns: 1,
* allowedTools: [],
* });
* console.log(result.text); // "Add user authentication"
* ```
*/
export async function simpleQuery(options: SimpleQueryOptions): Promise<SimpleQueryResult> {
const model = options.model || DEFAULT_MODEL;
const provider = ProviderFactory.getProviderForModel(model);
const bareModel = stripProviderPrefix(model);
let responseText = '';
let structuredOutput: Record<string, unknown> | undefined;
// Build provider options
const providerOptions = {
prompt: options.prompt,
model: bareModel,
originalModel: model,
cwd: options.cwd,
systemPrompt: options.systemPrompt,
maxTurns: options.maxTurns ?? 1,
allowedTools: options.allowedTools ?? [],
abortController: options.abortController,
outputFormat: options.outputFormat,
thinkingLevel: options.thinkingLevel,
reasoningEffort: options.reasoningEffort,
readOnly: options.readOnly,
settingSources: options.settingSources,
claudeApiProfile: options.claudeApiProfile, // Legacy: Pass active Claude API profile for alternative endpoint configuration
claudeCompatibleProvider: options.claudeCompatibleProvider, // New: Pass Claude-compatible provider (takes precedence)
credentials: options.credentials, // Pass credentials for resolving 'credentials' apiKeySource
};
for await (const msg of provider.executeQuery(providerOptions)) {
// Handle error messages
if (msg.type === 'error') {
const errorMessage = msg.error || 'Provider returned an error';
throw new Error(errorMessage);
}
// Extract text from assistant messages
if (msg.type === 'assistant' && msg.message?.content) {
for (const block of msg.message.content) {
if (block.type === 'text' && block.text) {
responseText += block.text;
}
}
}
// Handle result messages
if (msg.type === 'result') {
if (msg.subtype === 'success') {
// Use result text if longer than accumulated text
if (msg.result && msg.result.length > responseText.length) {
responseText = msg.result;
}
// Capture structured output if present
if (msg.structured_output) {
structuredOutput = msg.structured_output;
}
} else if (msg.subtype === 'error_max_turns') {
// Max turns reached - return what we have
break;
} else if (msg.subtype === 'error_max_structured_output_retries') {
throw new Error('Could not produce valid structured output after retries');
}
}
}
return { text: responseText, structured_output: structuredOutput };
}
/**
* Execute a streaming query with event callbacks
*
* Use this for queries where you need real-time progress updates,
* such as when displaying streaming output to a user.
*
* @example
* ```typescript
* const result = await streamingQuery({
* prompt: 'Analyze this project and suggest improvements',
* cwd: '/path/to/project',
* maxTurns: 250,
* allowedTools: ['Read', 'Glob', 'Grep'],
* onText: (text) => emitProgress(text),
* onToolUse: (tool, input) => emitToolUse(tool, input),
* });
* ```
*/
export async function streamingQuery(options: StreamingQueryOptions): Promise<SimpleQueryResult> {
const model = options.model || DEFAULT_MODEL;
const provider = ProviderFactory.getProviderForModel(model);
const bareModel = stripProviderPrefix(model);
let responseText = '';
let structuredOutput: Record<string, unknown> | undefined;
// Build provider options
const providerOptions = {
prompt: options.prompt,
model: bareModel,
originalModel: model,
cwd: options.cwd,
systemPrompt: options.systemPrompt,
maxTurns: options.maxTurns ?? 250,
allowedTools: options.allowedTools ?? ['Read', 'Glob', 'Grep'],
abortController: options.abortController,
outputFormat: options.outputFormat,
thinkingLevel: options.thinkingLevel,
reasoningEffort: options.reasoningEffort,
readOnly: options.readOnly,
settingSources: options.settingSources,
claudeApiProfile: options.claudeApiProfile, // Legacy: Pass active Claude API profile for alternative endpoint configuration
claudeCompatibleProvider: options.claudeCompatibleProvider, // New: Pass Claude-compatible provider (takes precedence)
credentials: options.credentials, // Pass credentials for resolving 'credentials' apiKeySource
};
for await (const msg of provider.executeQuery(providerOptions)) {
// Handle error messages
if (msg.type === 'error') {
const errorMessage = msg.error || 'Provider returned an error';
throw new Error(errorMessage);
}
// Extract content from assistant messages
if (msg.type === 'assistant' && msg.message?.content) {
for (const block of msg.message.content) {
if (block.type === 'text' && block.text) {
responseText += block.text;
options.onText?.(block.text);
} else if (block.type === 'tool_use' && block.name) {
options.onToolUse?.(block.name, block.input);
} else if (block.type === 'thinking' && block.thinking) {
options.onThinking?.(block.thinking);
}
}
}
// Handle result messages
if (msg.type === 'result') {
if (msg.subtype === 'success') {
// Use result text if longer than accumulated text
if (msg.result && msg.result.length > responseText.length) {
responseText = msg.result;
}
// Capture structured output if present
if (msg.structured_output) {
structuredOutput = msg.structured_output;
}
} else if (msg.subtype === 'error_max_turns') {
// Max turns reached - return what we have
break;
} else if (msg.subtype === 'error_max_structured_output_retries') {
throw new Error('Could not produce valid structured output after retries');
}
}
}
return { text: responseText, structured_output: structuredOutput };
}