mirror of
https://github.com/AutoMaker-Org/automaker.git
synced 2026-02-07 22:23:08 +00:00
feat: implement simple query service and enhance provider abstraction
- Introduced a new Simple Query Service to streamline basic AI queries, allowing for structured JSON outputs. - Updated existing routes to utilize the new service, replacing direct SDK calls with a unified interface for querying. - Enhanced provider handling in various routes, including generate-spec, generate-features-from-spec, and validate-issue, to support both Claude and Cursor models seamlessly. - Added structured output support for improved response handling and error management across the application.
This commit is contained in:
webdevcody
@@ -1,5 +1,5 @@
|
||||
/**
|
||||
* POST /validate-issue endpoint - Validate a GitHub issue using Claude SDK or Cursor (async)
|
||||
* POST /validate-issue endpoint - Validate a GitHub issue using provider abstraction (async)
|
||||
*
|
||||
* Scans the codebase to determine if an issue is valid, invalid, or needs clarification.
|
||||
* Runs asynchronously and emits events for progress and completion.
|
||||
@@ -7,7 +7,6 @@
|
||||
*/
|
||||
|
||||
import type { Request, Response } from 'express';
|
||||
import { query } from '@anthropic-ai/claude-agent-sdk';
|
||||
import type { EventEmitter } from '../../../lib/events.js';
|
||||
import type {
|
||||
IssueValidationResult,
|
||||
@@ -18,12 +17,11 @@ import type {
|
||||
LinkedPRInfo,
|
||||
ThinkingLevel,
|
||||
} from '@automaker/types';
|
||||
import { isCursorModel, DEFAULT_PHASE_MODELS, stripProviderPrefix } from '@automaker/types';
|
||||
import { isCursorModel, DEFAULT_PHASE_MODELS } from '@automaker/types';
|
||||
import { resolvePhaseModel } from '@automaker/model-resolver';
|
||||
import { createSuggestionsOptions } from '../../../lib/sdk-options.js';
|
||||
import { extractJson } from '../../../lib/json-extractor.js';
|
||||
import { writeValidation } from '../../../lib/validation-storage.js';
|
||||
import { ProviderFactory } from '../../../providers/provider-factory.js';
|
||||
import { streamingQuery } from '../../../providers/simple-query-service.js';
|
||||
import {
|
||||
issueValidationSchema,
|
||||
ISSUE_VALIDATION_SYSTEM_PROMPT,
|
||||
@@ -102,7 +100,7 @@ async function runValidation(
|
||||
|
||||
try {
|
||||
// Build the prompt (include comments and linked PRs if provided)
|
||||
const prompt = buildValidationPrompt(
|
||||
const basePrompt = buildValidationPrompt(
|
||||
issueNumber,
|
||||
issueTitle,
|
||||
issueBody,
|
||||
@@ -111,20 +109,15 @@ async function runValidation(
|
||||
linkedPRs
|
||||
);
|
||||
|
||||
let validationResult: IssueValidationResult | null = null;
|
||||
let responseText = '';
|
||||
|
||||
// Route to appropriate provider based on model
|
||||
if (isCursorModel(model)) {
|
||||
// Use Cursor provider for Cursor models
|
||||
logger.info(`Using Cursor provider for validation with model: ${model}`);
|
||||
// Determine if we should use structured output (Claude supports it, Cursor doesn't)
|
||||
const useStructuredOutput = !isCursorModel(model);
|
||||
|
||||
const provider = ProviderFactory.getProviderForModel(model);
|
||||
// Strip provider prefix - providers expect bare model IDs
|
||||
const bareModel = stripProviderPrefix(model);
|
||||
|
||||
// For Cursor, include the system prompt and schema in the user prompt
|
||||
const cursorPrompt = `${ISSUE_VALIDATION_SYSTEM_PROMPT}
|
||||
// Build the final prompt - for Cursor, include system prompt and JSON schema instructions
|
||||
let finalPrompt = basePrompt;
|
||||
if (!useStructuredOutput) {
|
||||
finalPrompt = `${ISSUE_VALIDATION_SYSTEM_PROMPT}
|
||||
|
||||
CRITICAL INSTRUCTIONS:
|
||||
1. DO NOT write any files. Return the JSON in your response only.
|
||||
@@ -135,121 +128,71 @@ ${JSON.stringify(issueValidationSchema, null, 2)}
|
||||
|
||||
Your entire response should be valid JSON starting with { and ending with }. No text before or after.
|
||||
|
||||
${prompt}`;
|
||||
|
||||
for await (const msg of provider.executeQuery({
|
||||
prompt: cursorPrompt,
|
||||
model: bareModel,
|
||||
cwd: projectPath,
|
||||
readOnly: true, // Issue validation only reads code, doesn't write
|
||||
})) {
|
||||
if (msg.type === 'assistant' && msg.message?.content) {
|
||||
for (const block of msg.message.content) {
|
||||
if (block.type === 'text' && block.text) {
|
||||
responseText += block.text;
|
||||
|
||||
// Emit progress event
|
||||
const progressEvent: IssueValidationEvent = {
|
||||
type: 'issue_validation_progress',
|
||||
issueNumber,
|
||||
content: block.text,
|
||||
projectPath,
|
||||
};
|
||||
events.emit('issue-validation:event', progressEvent);
|
||||
}
|
||||
}
|
||||
} else if (msg.type === 'result' && msg.subtype === 'success' && msg.result) {
|
||||
// Use result if it's a final accumulated message
|
||||
if (msg.result.length > responseText.length) {
|
||||
responseText = msg.result;
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// Parse JSON from the response text using shared utility
|
||||
if (responseText) {
|
||||
validationResult = extractJson<IssueValidationResult>(responseText, { logger });
|
||||
}
|
||||
} else {
|
||||
// Use Claude SDK for Claude models
|
||||
logger.info(`Using Claude provider for validation with model: ${model}`);
|
||||
|
||||
// Load autoLoadClaudeMd setting
|
||||
const autoLoadClaudeMd = await getAutoLoadClaudeMdSetting(
|
||||
projectPath,
|
||||
settingsService,
|
||||
'[ValidateIssue]'
|
||||
);
|
||||
|
||||
// Use thinkingLevel from request if provided, otherwise fall back to settings
|
||||
let effectiveThinkingLevel: ThinkingLevel | undefined = thinkingLevel;
|
||||
if (!effectiveThinkingLevel) {
|
||||
const settings = await settingsService?.getGlobalSettings();
|
||||
const phaseModelEntry =
|
||||
settings?.phaseModels?.validationModel || DEFAULT_PHASE_MODELS.validationModel;
|
||||
const resolved = resolvePhaseModel(phaseModelEntry);
|
||||
effectiveThinkingLevel = resolved.thinkingLevel;
|
||||
}
|
||||
|
||||
// Create SDK options with structured output and abort controller
|
||||
const options = createSuggestionsOptions({
|
||||
cwd: projectPath,
|
||||
model: model as ModelAlias,
|
||||
systemPrompt: ISSUE_VALIDATION_SYSTEM_PROMPT,
|
||||
abortController,
|
||||
autoLoadClaudeMd,
|
||||
thinkingLevel: effectiveThinkingLevel,
|
||||
outputFormat: {
|
||||
type: 'json_schema',
|
||||
schema: issueValidationSchema as Record<string, unknown>,
|
||||
},
|
||||
});
|
||||
|
||||
// Execute the query
|
||||
const stream = query({ prompt, options });
|
||||
|
||||
for await (const msg of stream) {
|
||||
// Collect assistant text for debugging and emit progress
|
||||
if (msg.type === 'assistant' && msg.message?.content) {
|
||||
for (const block of msg.message.content) {
|
||||
if (block.type === 'text') {
|
||||
responseText += block.text;
|
||||
|
||||
// Emit progress event
|
||||
const progressEvent: IssueValidationEvent = {
|
||||
type: 'issue_validation_progress',
|
||||
issueNumber,
|
||||
content: block.text,
|
||||
projectPath,
|
||||
};
|
||||
events.emit('issue-validation:event', progressEvent);
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
// Extract structured output on success
|
||||
if (msg.type === 'result' && msg.subtype === 'success') {
|
||||
const resultMsg = msg as { structured_output?: IssueValidationResult };
|
||||
if (resultMsg.structured_output) {
|
||||
validationResult = resultMsg.structured_output;
|
||||
logger.debug('Received structured output:', validationResult);
|
||||
}
|
||||
}
|
||||
|
||||
// Handle errors
|
||||
if (msg.type === 'result') {
|
||||
const resultMsg = msg as { subtype?: string };
|
||||
if (resultMsg.subtype === 'error_max_structured_output_retries') {
|
||||
logger.error('Failed to produce valid structured output after retries');
|
||||
throw new Error('Could not produce valid validation output');
|
||||
}
|
||||
}
|
||||
}
|
||||
${basePrompt}`;
|
||||
}
|
||||
|
||||
// Load autoLoadClaudeMd setting
|
||||
const autoLoadClaudeMd = await getAutoLoadClaudeMdSetting(
|
||||
projectPath,
|
||||
settingsService,
|
||||
'[ValidateIssue]'
|
||||
);
|
||||
|
||||
// Use thinkingLevel from request if provided, otherwise fall back to settings
|
||||
let effectiveThinkingLevel: ThinkingLevel | undefined = thinkingLevel;
|
||||
if (!effectiveThinkingLevel) {
|
||||
const settings = await settingsService?.getGlobalSettings();
|
||||
const phaseModelEntry =
|
||||
settings?.phaseModels?.validationModel || DEFAULT_PHASE_MODELS.validationModel;
|
||||
const resolved = resolvePhaseModel(phaseModelEntry);
|
||||
effectiveThinkingLevel = resolved.thinkingLevel;
|
||||
}
|
||||
|
||||
logger.info(`Using model: ${model}`);
|
||||
|
||||
// Use streamingQuery with event callbacks
|
||||
const result = await streamingQuery({
|
||||
prompt: finalPrompt,
|
||||
model: model as string,
|
||||
cwd: projectPath,
|
||||
systemPrompt: useStructuredOutput ? ISSUE_VALIDATION_SYSTEM_PROMPT : undefined,
|
||||
abortController,
|
||||
thinkingLevel: effectiveThinkingLevel,
|
||||
readOnly: true, // Issue validation only reads code, doesn't write
|
||||
settingSources: autoLoadClaudeMd ? ['user', 'project', 'local'] : undefined,
|
||||
outputFormat: useStructuredOutput
|
||||
? {
|
||||
type: 'json_schema',
|
||||
schema: issueValidationSchema as Record<string, unknown>,
|
||||
}
|
||||
: undefined,
|
||||
onText: (text) => {
|
||||
responseText += text;
|
||||
// Emit progress event
|
||||
const progressEvent: IssueValidationEvent = {
|
||||
type: 'issue_validation_progress',
|
||||
issueNumber,
|
||||
content: text,
|
||||
projectPath,
|
||||
};
|
||||
events.emit('issue-validation:event', progressEvent);
|
||||
},
|
||||
});
|
||||
|
||||
// Clear timeout
|
||||
clearTimeout(timeoutId);
|
||||
|
||||
// Get validation result from structured output or parse from text
|
||||
let validationResult: IssueValidationResult | null = null;
|
||||
|
||||
if (result.structured_output) {
|
||||
validationResult = result.structured_output as unknown as IssueValidationResult;
|
||||
logger.debug('Received structured output:', validationResult);
|
||||
} else if (responseText) {
|
||||
// Parse JSON from response text
|
||||
validationResult = extractJson<IssueValidationResult>(responseText, { logger });
|
||||
}
|
||||
|
||||
// Require validation result
|
||||
if (!validationResult) {
|
||||
logger.error('No validation result received from AI provider');
|
||||
@@ -299,7 +242,7 @@ ${prompt}`;
|
||||
/**
|
||||
* Creates the handler for validating GitHub issues against the codebase.
|
||||
*
|
||||
* Uses Claude SDK with:
|
||||
* Uses the provider abstraction with:
|
||||
* - Read-only tools (Read, Glob, Grep) for codebase analysis
|
||||
* - JSON schema structured output for reliable parsing
|
||||
* - System prompt guiding the validation process
|
||||
|
||||
Reference in New Issue
Block a user