EN KO

AI-Powered Analysis

Connect Noir to Large Language Models (cloud-based, local, or ACP agent-based) for deeper code analysis. AI helps identify endpoints in unsupported languages and frameworks.

Usage

Specify an AI provider, model, and API key:

noir -b . --ai-provider <PROVIDER> --ai-model <MODEL_NAME> --ai-key <YOUR_API_KEY>

For ACP providers (acp:*), --ai-model is optional and --ai-key is usually not required:

noir -b . --ai-provider acp:codex

Command-Line Flags

Flag Description
--ai-provider Provider prefix (e.g., openai, ollama, acp:codex) or custom API URL
--ai-model Model name (e.g., gpt-4o), optional for acp:*
--ai-key API key (or use NOIR_AI_KEY env var)
--ai-agent Enable agentic AI workflow (iterative tool-calling loop)
--ai-agent-max-steps Max steps for AI agent loop (default: 20)
--ai-native-tools-allowlist Provider allowlist for native tool-calling (comma-separated, default: openai,xai,github)
--ai-max-token Max tokens for AI requests (optional)
--cache-disable Disable LLM response cache
--cache-clear Clear LLM cache before run

Supported AI Providers

Noir has built-in presets for several popular AI providers:

Prefix Default Host
openai https://api.openai.com
xai https://api.x.ai
github https://models.github.ai
azure https://models.inference.ai.azure.com
openrouter https://openrouter.ai/api/v1
vllm http://localhost:8000
ollama http://localhost:11434
lmstudio http://localhost:1234
acp:codex npx @zed-industries/codex-acp
acp:gemini gemini --experimental-acp
acp:claude npx @zed-industries/claude-agent-acp

For custom providers, use the full API URL: --ai-provider=http://my-custom-api:9000.

For raw ACP and agent stderr logs, set NOIR_ACP_RAW_LOG=1.

How AI-Powered Analysis Works

flowchart TB
    Start([Start AI Analysis]) --> InitAdapter[Initialize LLM Adapter]
    InitAdapter --> ProviderCheck{Provider Type?}
    
    ProviderCheck -->|OpenAI/xAI/etc| GeneralAdapter[General Adapter
OpenAI-compatible API]
    ProviderCheck -->|Ollama/Local| OllamaAdapter[Ollama Adapter
with Context Reuse]
    ProviderCheck -->|ACP Agent| ACPAdapter[ACP Adapter
Codex/Gemini/Claude/Custom]
    
    GeneralAdapter --> FileSelection
    OllamaAdapter --> FileSelection
    ACPAdapter --> FileSelection
    
    FileSelection[File Selection] --> FileCount{File Count?}
    
    FileCount -->|≤ 10 files| AnalyzeAll[Analyze All Files]
    FileCount -->|> 10 files| LLMFilter[LLM-Based Filtering]
    
    LLMFilter --> CacheCheck1{Cache Hit?}
    CacheCheck1 -->|Yes| UseCached1[Use Cached Filter]
    CacheCheck1 -->|No| FilterLLM[Call LLM with
FILTER prompt]
    FilterLLM --> StoreCache1[Store in Cache]
    UseCached1 --> TargetFiles
    StoreCache1 --> TargetFiles
    
    TargetFiles[Selected Target Files] --> BundleCheck{Large File Set
and Token Limit?}
    
    AnalyzeAll --> BundleCheck
    
    BundleCheck -->|Yes| BundleMode[Bundle Analysis Mode]
    BundleCheck -->|No| SingleMode[Single File Mode]
    
    BundleMode --> CreateBundles[Create File Bundles
within Token Limits]
    CreateBundles --> ParallelBundles[Process Bundles
Concurrently]
    
    ParallelBundles --> BundleLoop{For Each Bundle}
    BundleLoop --> CacheCheck2{Cache Hit?}
    CacheCheck2 -->|Yes| UseCached2[Use Cached Analysis]
    CacheCheck2 -->|No| BundleLLM[Call LLM with
BUNDLE_ANALYZE prompt]
    BundleLLM --> StoreCache2[Store in Cache]
    UseCached2 --> ParseEndpoints1
    StoreCache2 --> ParseEndpoints1
    ParseEndpoints1[Parse Endpoints
from Response] --> BundleLoop
    BundleLoop -->|Done| Combine
    
    SingleMode --> FileLoop{For Each File}
    FileLoop --> CacheCheck3{Cache Hit?}
    CacheCheck3 -->|Yes| UseCached3[Use Cached Analysis]
    CacheCheck3 -->|No| AnalyzeLLM[Call LLM with
ANALYZE prompt]
    AnalyzeLLM --> StoreCache3[Store in Cache]
    UseCached3 --> ParseEndpoints2
    StoreCache3 --> ParseEndpoints2
    ParseEndpoints2[Parse Endpoints
from Response] --> FileLoop
    FileLoop -->|Done| Combine
    
    Combine[Combine All Endpoints] --> LLMOptCheck{LLM Optimization
Enabled?}
    
    LLMOptCheck -->|Yes| FindCandidates[Find Optimization
Candidates]
    FindCandidates --> OptLoop{For Each Candidate}
    OptLoop --> OptimizeLLM[Call LLM with
OPTIMIZE prompt]
    OptimizeLLM --> ApplyOpt[Apply Optimizations
to Endpoint]
    ApplyOpt --> OptLoop
    OptLoop -->|Done| FinalResults
    
    LLMOptCheck -->|No| FinalResults[Final Optimized Results]
    
    FinalResults --> End([End])
    
    style Start fill:#e1f5e1
    style End fill:#e1f5e1
    style LLMFilter fill:#fff4e1
    style FilterLLM fill:#e1f0ff
    style BundleLLM fill:#e1f0ff
    style AnalyzeLLM fill:#e1f0ff
    style OptimizeLLM fill:#e1f0ff
    style CacheCheck1 fill:#ffe1e1
    style CacheCheck2 fill:#ffe1e1
    style CacheCheck3 fill:#ffe1e1
    style UseCached1 fill:#e1ffe1
    style UseCached2 fill:#e1ffe1
    style UseCached3 fill:#e1ffe1

Key Components

LLM Adapter Layer

Provider-agnostic adapters: General (OpenAI-compatible APIs), Ollama (with server-side context reuse), and ACP (agent runtimes like acp:codex).

Intelligent File Filtering

For projects with >10 files, the LLM filters the file list to identify likely endpoint files before analysis.

Bundle Analysis

Groups files into token-limited bundles and processes them concurrently to maximize throughput on large codebases.

Response Caching

LLM responses are cached on disk (SHA256-keyed) at ~/.local/share/noir/cache/ai/. Use --cache-disable or --cache-clear to control caching.

LLM Optimizer

Optional post-processing that normalizes URLs, parameter names, and applies RESTful conventions to improve endpoint quality.

Esc