Skip to content

API.md

Latest commit

 

History

History
538 lines (444 loc) · 14.1 KB

API.md

File metadata and controls

538 lines (444 loc) · 14.1 KB
Raw

API Reference Guide — ARIA v1.5.0

This document details the REST endpoints and Server-Sent Event (SSE) streams exposed by the FastAPI backend server. The API Layer is part of the Repository Intelligence Architecture (RIA), the platform's modular production architecture.

All endpoints are versioned under the /api/v1 prefix. Legacy root paths (e.g. /api/...) are supported as backward-compatible shims.

Authentication

When the application is configured with an API_KEY (via the environment variable), all resource-intensive endpoints require authentication.

You must supply the API key in one of the following HTTP headers:

  1. X-API-Key Header:

    X-API-Key: your_secret_api_key_here

  2. Authorization Bearer Token Header:

    Authorization: Bearer your_secret_api_key_here

Failed authentication requests return a 401 Unauthorized response:

{
  "detail": "Unauthorized. Invalid or missing API key."
}

1. Repository Ingestion & Analysis

Index Repository (Sync)

Triggers synchronous metadata ingestion for a repository.

  • Endpoint: POST /api/v1/index or POST /api/index
  • Method: POST
  • Request Body: 
    {
  "repo_url": "https://github.com/fastapi/fastapi",
  "branch": "master"
}
  • Response (200 OK): 
    {
  "status": "completed",
  "owner": "fastapi",
  "repo_name": "fastapi",
  "files_count": 342,
  "symbols_count": 1821
}

Index Repository (Streaming Analysis)

Triggers repository cloning and AST parsing, streaming progress updates as Server-Sent Events (SSE).

  • Endpoint: POST /api/v1/analyze or POST /api/analyze
  • Method: POST
  • Request Body: 
    {
  "url": "https://github.com/fastapi/fastapi",
  "branch": "master",
  "model": "deepseek-ai/deepseek-v4-flash",
  "force_rebuild": false
}
  • SSE Stream Data (Progress Events): 
    event: progress
data: {"status": "cloning", "percent": 15, "message": "Cloning repository..."}

event: progress
data: {"status": "parsing", "percent": 50, "message": "Parsing abstract syntax trees..."}

event: progress
data: {"status": "completed", "percent": 100, "message": "Analysis completed."}

Get Ingested Analysis Details

Returns metadata for an already analyzed repository.

  • Endpoint: GET /api/v1/analysis/{owner}/{repo_name} or GET /api/analysis/{owner}/{repo_name}
  • Method: GET
  • Response (200 OK): 
    {
  "owner": "fastapi",
  "repo_name": "fastapi",
  "indexed_at": "2026-07-04T12:00:00Z",
  "primary_language": "Python",
  "files_indexed": 342,
  "status": "ready"
}

Repair Repository Index

Rebuilds the dependency graph and symbol index from the already-cloned repository on disk without re-cloning or re-embedding.

  • Endpoint: POST /api/v1/repos/repair or POST /api/repos/repair
  • Method: POST
  • Request Body: 
    {
  "owner": "fastapi",
  "repo": "fastapi"
}
  • Response (200 OK): 
    {
  "status": "completed",
  "message": "Repository index repaired successfully."
}

2. Codebase Chat & Issue Engine

Chat Query (Streaming)

Queries the multi-agent chatbot about the codebase. Responses are streamed as SSE chunks.

  • Endpoint: POST /api/v1/chat or POST /api/chat
  • Method: POST
  • Request Body: 
    {
  "repo": "fastapi/fastapi",
  "message": "What is the entry point of the app?",
  "history": [
    {"role": "user", "content": "Hi"},
    {"role": "assistant", "content": "Hello! How can I help?"}
  ]
}
  • Headers Required: Accept: text/event-stream
  • SSE Stream Data Chunks: 
    event: chunk
data: "The "

event: chunk
data: "main "

event: chunk
data: "entrypoint "

event: citations
data: [{"file": "fastapi/main.py", "lines": "1-15", "confidence": 0.95}]

event: done
data: [DONE]

Chat Health & LLM Provider Status

Runs a live health check against every configured LLM provider.

  • Endpoint: GET /api/v1/chat/health or GET /api/chat/health
  • Method: GET
  • Response (200 OK): 
    {
  "status": "ok",
  "healthy": true,
  "primary_provider": "gemini",
  "healthy_providers": ["gemini", "deepseek"],
  "unhealthy_providers": []
}

Reload LLM Providers

Hot-reloads the LLM provider configuration from .env without restarting the server.

  • Endpoint: POST /api/v1/chat/reload or POST /api/chat/reload
  • Method: POST
  • Response (200 OK): 
    {
  "status": "success",
  "message": "LLM provider reloaded successfully. Chat is ready."
}

Map GitHub Issue to Implementation Plan

Generates a grounded implementation plan for a given issue using a two-LLM-call pipeline.

  • Endpoint: POST /api/v1/issues/map or POST /api/issues/map
  • Method: POST
  • Request Body: 
    {
  "repo": "fastapi/fastapi",
  "title": "Fix memory leaks in cache",
  "description": "The cache store is leaking reference counts."
}
  • Response (200 OK): 
    {
  "issue_summary": "Fix memory leaks in cache",
  "issue_type": "Bugfix",
  "relevant_files": ["backend/dependencies.py"],
  "affected_components": ["Cache System"],
  "implementation_plan": [
    {
      "step": 1,
      "action": "Introduce eviction logic to clean stale references."
    }
  ],
  "complexity": "Medium",
  "confidence": 0.9,
  "verified": true,
  "sources": ["backend/dependencies.py"]
}

3. Visualizations, Graphs, and Metrics

Get File Dependency Graph

Returns node and edge representations for the file-level import graph.

  • Endpoint: GET /api/v1/graph/{owner}/{repo}/full or GET /api/graph/{owner}/{repo}/full
  • Method: GET
  • Query Parameters:
    • q: Optional search keyword to filter nodes.
  • Response (200 OK): 
    {
  "nodes": [
    {
      "id": "fastapi/main.py",
      "label": "main.py",
      "category": "entry_point",
      "highlighted": false,
      "is_focus": false
    }
  ],
  "edges": [
    {
      "source": "fastapi/main.py",
      "target": "fastapi/applications.py",
      "relationship": "imports"
    }
  ],
  "matched_count": 1
}

Get Neighborhood Subgraph

Returns only the immediate imports and dependents of a focus node.

  • Endpoint: GET /api/v1/graph/{owner}/{repo}/neighbors/{focus_id} or GET /api/graph/{owner}/{repo}/neighbors/{focus_id}
  • Method: GET
  • Response (200 OK): (Same schema format as full graph, filtered to neighborhood nodes)

Trace Paths (Reachability/Blast Radius)

Computes a reachability subgraph using BFS in forward, backward, or bidirectional orientations.

  • Endpoint: GET /api/v1/graph/{owner}/{repo}/trace/{focus_id} or GET /api/graph/{owner}/{repo}/trace/{focus_id}
  • Method: GET
  • Query Parameters:
    • direction: forward (imports), backward (dependents), or both.
    • depth: Maximum depth search limit (default 6).
  • Response (200 OK): (Returns sub-graph representing reachability paths)

Get Call Graph Summary

Returns function-level call graph details for the repository.

  • Endpoint: GET /api/v1/call-graph/{owner}/{repo} or GET /api/call-graph/{owner}/{repo}
  • Method: GET
  • Response (200 OK): (Returns function node call mappings)

Get Function Blast Radius

Computes the blast radius of changing a specific function.

  • Endpoint: GET /api/v1/call-graph/{owner}/{repo}/blast-radius/{function_id} or GET /api/call-graph/{owner}/{repo}/blast-radius/{function_id}
  • Method: GET
  • Response (200 OK): 
    {
  "function_id": "fastapi/main.py:app",
  "affected_functions": ["fastapi/applications.py:FastAPI"],
  "affected_files": ["fastapi/applications.py"],
  "depth": 1,
  "risk_level": "LOW",
  "recursive_cycles": []
}

Get Churn Hotspots

Identifies code files with high commit frequency and high topological dependency centrality.

  • Endpoint: GET /api/v1/churn/{owner}/{repo}/hotspots or GET /api/churn/{owner}/{repo}/hotspots
  • Method: GET
  • Query Parameters:
    • top_n: Maximum items to return (default 25).
    • since_days: Git history window (default 365).
  • Response (200 OK): 
    {
  "hotspots": [
    {
      "file_path": "backend/api.py",
      "commit_count": 42,
      "churn_score": 8.4
    }
  ]
}

4. Reports & Engineering Advisor

Compile Health Report

Triggers report generation or fetches the cached analysis report.

  • Endpoint: POST /api/v1/report/{owner}/{repo}/build
  • Method: POST
  • Response (200 OK): 
    {
  "metadata": {
    "repo_name": "fastapi",
    "owner": "fastapi",
    "total_loc": 25410,
    "generated_at": "2026-07-04T18:00:00Z"
  },
  "scores": {
    "overall": 88,
    "architecture": 90,
    "api": 85,
    "hygiene": 92,
    "churn": 80,
    "readability": 95,
    "grade": "A"
  },
  "refactoring_priorities": [
    "Refactor volatile hotspot module: fastapi/applications.py (churn score: 95.0)"
  ]
}

Download Report Files

Downloads the report in static formats.

  • Endpoint: GET /api/v1/report/{owner}/{repo}/download
  • Method: GET
  • Query Parameters:
    • format: html, markdown, or pdf.
  • Response: File attachment stream (text/html, text/markdown, or application/pdf).

Generate AI Engineering Advisor Report

Analyzes the digital twin structure and continuous monitoring states to generate prioritized advisor recommendations and roadmaps.

  • Endpoint: POST /api/v1/repositories/{owner}/{repo}/advisor or POST /api/repositories/{owner}/{repo}/advisor
  • Method: POST
  • Response (200 OK): 
    {
  "repository": "fastapi/fastapi",
  "overall_priority": "medium",
  "total_recommendations": 12,
  "top_recommendations": [
    {
      "id": "rec-1",
      "title": "Resolve cycle between main and routing",
      "category": "architecture",
      "priority": "high",
      "estimated_effort": "medium"
    }
  ],
  "roadmap_phases": 3,
  "roadmap_summary": [
    {
      "phase": 1,
      "title": "Mitigate Security & Dependency Issues",
      "recommendation_count": 4
    }
  ]
}

Get Latest Advisor Report

Fetches the most recently generated advisor report details.

  • Endpoint: GET /api/v1/repositories/{owner}/{repo}/advisor/latest or GET /api/repositories/{owner}/{repo}/advisor/latest
  • Method: GET
  • Response (200 OK): (Same schema format as Advisor Report)

Get Engineering Roadmap phases

Fetches the roadmap list from the latest Advisor report.

  • Endpoint: GET /api/v1/repositories/{owner}/{repo}/advisor/roadmap or GET /api/repositories/{owner}/{repo}/advisor/roadmap
  • Method: GET
  • Response (200 OK): (Returns Roadmap Phases array)

5. Execution Planner (AEA²)

Generate Execution Plan

Constructs an autonomous implementation roadmap from the latest Advisor report.

  • Endpoint: POST /api/v1/repositories/{owner}/{repo}/execution-plan or POST /api/repositories/{owner}/{repo}/execution-plan
  • Method: POST
  • Response (200 OK): 
    {
  "repository": "fastapi/fastapi",
  "total_tasks": 8,
  "total_batches": 2,
  "critical_path_length": 5,
  "rollback_checkpoints": 2,
  "conflict_count": 0,
  "overall_risk": "low",
  "batches": [
    {
      "batch_id": "batch-1",
      "order": 1,
      "title": "Initial Refactoring Pass",
      "task_count": 3,
      "parallel": true,
      "estimated_effort": "1 day"
    }
  ],
  "critical_path": ["task-1", "task-3"],
  "metadata": {}
}

Get Latest Execution Plan

Fetches the most recently generated execution plan.

  • Endpoint: GET /api/v1/repositories/{owner}/{repo}/execution-plan/latest or GET /api/repositories/{owner}/{repo}/execution-plan/latest
  • Method: GET
  • Response (200 OK): (Same schema format as Execution Plan)

6. Intelligent IDE Workspace snapshot

Get Full Workspace Snapshot

Returns the consolidated data for all workspace views to power the IDE dashboard panels in a single request.

  • Endpoint: GET /api/v1/repositories/{owner}/{repo}/workspace or GET /api/repositories/{owner}/{repo}/workspace
  • Method: GET
  • Query Parameters:
    • file: Optional path of currently open file to contextualize snapshot.
    • symbol: Optional currently active symbol.
    • panel: Active panel focus name.
  • Response (200 OK): (Returns WorkspaceSnapshot containing Overview, Explorer, Chat, Findings, Timeline, Monitor, Advisor, and Execution panel structures)

Get Workspace Panels Individually

Retrieve individual panel snapshots using these sub-routes (available with root, /api, and /api/v1 prefixes):

  • GET /api/v1/repositories/{owner}/{repo}/workspace/overview (Overview Panel)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/explorer (Knowledge Graph Explorer)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/findings (Findings Panel)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/timeline (Timeline Panel)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/monitor (Continuous Monitor Panel)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/advisor (Advisor Panel)
  • GET /api/v1/repositories/{owner}/{repo}/workspace/execution (Execution Plan Panel)

7. Health, Status, and Observability

Server Live Status

  • Endpoint: GET /health or GET /api/v1/health
  • Method: GET
  • Response (200 OK): 
    {
  "status": "healthy",
  "backend": "online",
  "llm_provider": "gemini",
  "llm_model": "gemini-2.5-flash",
  "embedding_provider": "BAAI/bge-small-en-v1.5",
  "vector_db": "chromadb"
}

Prometheus Metrics

  • Endpoint: GET /metrics or GET /api/v1/metrics
  • Method: GET
  • Response (200 OK): Standard Prometheus text format listing HTTP request counts, active gauges, and processing latencies.