Files
MtoRagSystem/config/services.yaml

119 lines
3.8 KiB
YAML
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# ------------------------------------------------------------
# Parameters
# ------------------------------------------------------------
parameters:
mto.index.chunk_size: 800
mto.index.chunk_overlap: 100
mto.index.embedding_model: 'nomic-embed-text'
mto.index.embedding_dimension: 768
mto.index.scoring_version: 1
mto.vector.python_bin: '/var/www/html/.venv/bin/python3'
mto.vector.ingest_script: '/src/Vector/vector_ingest.py'
mto.vector.timeout: 600
# ------------------------------------------------------------
# Services
# ------------------------------------------------------------
services:
# ------------------------------------------------------------
# Default service configuration
# ------------------------------------------------------------
_defaults:
autowire: true
autoconfigure: true
bind:
Psr\Log\LoggerInterface $agentLogger: '@monolog.logger.agent'
string $projectDir: '%kernel.project_dir%'
# ------------------------------------------------------------
# Automatically register all services in src/
# ------------------------------------------------------------
App\:
resource: '../src/'
exclude:
- '../src/DependencyInjection/'
- '../src/Entity/'
- '../src/Kernel.php'
App\Controller\:
resource: '../src/Controller/'
tags: ['controller.service_arguments']
# ------------------------------------------------------------
# AI Agent Infrastructure
# ------------------------------------------------------------
App\Infrastructure\OllamaClient:
arguments:
$apiUrl: '%env(AI_LLM_API_URL)%'
$model: '%env(AI_LLM_MODEL)%'
$timeoutSeconds: '%env(int:AI_LLM_TIMEOUT)%'
# ------------------------------------------------------------
# AI Agent Context & Runner
# ------------------------------------------------------------
App\Context\ContextService:
arguments:
$historyDir: '%env(AI_HISTORY_DIR)%'
$projectDir: '%kernel.project_dir%'
App\Agent\AgentRunner:
arguments:
$debug: '%env(bool:AI_DEBUG)%'
$logPrompt: '%env(bool:AI_LOG_PROMPT)%'
$logContext: '%env(bool:AI_LOG_CONTEXT)%'
# ------------------------------------------------------------
# NDJSON Retrieval Stack (FINAL ARCHITECTURE)
# ------------------------------------------------------------
App\Knowledge\Retrieval\NdjsonChunkLookup: ~
App\Knowledge\Retrieval\NdjsonKeywordSearch: ~
App\Knowledge\Retrieval\NdjsonHybridRetriever:
arguments:
$maxChunks: 3
$vectorTopK: 5
App\Knowledge\Retrieval\CachedRetriever:
arguments:
$inner: '@App\Knowledge\Retrieval\NdjsonHybridRetriever'
$cache: '@cache.app'
$ttlSeconds: 600
App\Knowledge\Retrieval\RetrieverInterface:
alias: App\Knowledge\Retrieval\CachedRetriever
# ------------------------------------------------------------
# Vector Search (FAISS NDJSON-based)
# ------------------------------------------------------------
App\Vector\VectorSearchClient:
arguments:
$vectorDir: '%kernel.project_dir%/var/knowledge'
App\Vector\VectorIndexBuilder:
arguments:
$pythonBin: '%mto.vector.python_bin%'
$relativeScriptPath: '%mto.vector.ingest_script%'
$timeoutSeconds: '%mto.vector.timeout%'
# ------------------------------------------------------------
# Index Configuration (Guardrails)
# ------------------------------------------------------------
App\Index\IndexConfiguration:
arguments:
$chunkSize: '%mto.index.chunk_size%'
$chunkOverlap: '%mto.index.chunk_overlap%'
$embeddingModel: '%mto.index.embedding_model%'
$embeddingDimension: '%mto.index.embedding_dimension%'
$scoringVersion: '%mto.index.scoring_version%'
$indexFormat: 'ndjson'
$vectorBackend: 'faiss'