haiku.rag/docs/agents.md

Agents

Four agentic flows are provided by haiku.rag:

• Simple QA Agent — a focused question answering agent
• Chat Agent — multi-turn conversational RAG with session memory
• Research Graph — a multi-step research workflow with question decomposition
• RLM Agent — complex analytical tasks via sandboxed Python code execution (see RLM Agent)

See QA and Research Configuration for configuring model, iterations, concurrency, and other settings.

Simple QA Agent

The simple QA agent answers a single question using the knowledge base. It retrieves relevant chunks, optionally expands context around them, and asks the model to answer strictly based on that context.

Key points:

• Uses a single search_documents tool to fetch relevant chunks
• Can be run with or without inline citations in the prompt
• Returns a plain string answer

CLI usage:

haiku-rag ask "What is climate change?"

# With citations
haiku-rag ask "What is climate change?" --cite

# Deep mode (uses research graph with optimized settings)
haiku-rag ask "What are the main features of haiku.rag?" --deep

Python usage:

from haiku.rag.client import HaikuRAG
from haiku.rag.agents.qa.agent import QuestionAnswerAgent

async with HaikuRAG(path_to_db) as client:
    agent = QuestionAnswerAgent(
        client=client,
        provider="openai",
        model="gpt-4o-mini",
        use_citations=False,
    )

    answer = await agent.answer("What is climate change?")
    print(answer)

Chat Agent

The chat agent enables multi-turn conversational RAG. It maintains session state including Q/A history and uses that context to improve follow-up answers.

Key features:

• Session memory: Previous Q/A pairs are used as context for follow-up questions
• Query expansion: Search toolset generates multiple query variations for better recall
• Document filtering: Natural language document filtering ("search in document X about...")
• Confidence filtering: Low-confidence answers are flagged

Tools

The chat agent uses five tools:

• list_documents — Browse available documents in the knowledge base
• summarize_document — Generate a summary of a specific document
• get_document — Retrieve a specific document by title or URI
• search — Hybrid search with optional document filter
• ask — Answer questions using the conversational research graph (automatically recalls prior answers)

The ask tool automatically checks conversation history before running research. It uses embedding similarity (0.7 cosine threshold) to find semantically matching prior answers, which are passed to the research planner as context. When prior answers are sufficient, the planner can skip searching entirely.

CLI Usage

haiku-rag chat
haiku-rag chat --db /path/to/database.lancedb

See Applications for the full TUI interface guide.

Python Usage

from haiku.rag.client import HaikuRAG
from haiku.rag.agents.chat import create_chat_agent, ChatDeps
from haiku.rag.tools import ToolContext

async with HaikuRAG(path_to_db) as client:
    # Create agent with composed toolsets
    context = ToolContext()
    agent = create_chat_agent(config, client, context)
    deps = ChatDeps(config=config, tool_context=context)

    # First question
    result = await agent.run("What is haiku.rag?", deps=deps)
    print(result.output)

    # Follow-up (uses session context)
    result = await agent.run("How does it handle PDFs?", deps=deps)
    print(result.output)

Session State

The ChatSessionState maintains:

• session_id — Unique identifier for the session
• qa_history — List of previous Q/A pairs (FIFO, max 50)
• session_context — Automatically maintained session context summary
• document_filter — List of document titles/URIs to restrict searches
• citation_registry — Stable mapping of chunk IDs to citation indices

Citation Registry: Citation indices persist across tool calls within a session. The same chunk_id always returns the same citation index (first-occurrence-wins). This ensures consistent citation numbering in multi-turn conversations — [1] always refers to the same source.

# Example: citation indices are stable across calls
state = ChatSessionState()

# First call returns citations [1], [2], [3]
# Second call reuses [1] if same chunk, assigns [4], [5] for new chunks
# User can reference [1] in follow-up and it still refers to original source

Q/A history is used to:

1. Provide context for follow-up questions
2. Avoid repeating previous answers (the ask tool automatically recalls relevant prior answers)
3. Enable semantic ranking of relevant past answers

AG-UI Integration

When using the chat agent with AG-UI streaming, state is emitted under a namespaced key to avoid conflicts with other agents:

from haiku.rag.agents.chat import AGUI_STATE_KEY, ChatDeps
from haiku.rag.tools import ToolContext

# AGUI_STATE_KEY = "haiku.rag.chat"

context = ToolContext()
agent = create_chat_agent(config, client, context)
deps = ChatDeps(
    config=config,
    tool_context=context,
    state_key=AGUI_STATE_KEY,  # Enables namespaced state emission
)

The emitted state structure:

{
  "haiku.rag.chat": {
    "session_id": "",
    "citations": [...],
    "qa_history": [...],
    "document_filter": [...],
    "citation_registry": {"chunk-id-1": 1, "chunk-id-2": 2}
  }
}

Frontend clients should extract state from under this key. See the Web Application for a complete implementation example.

Research Graph

The research workflow is implemented as a typed pydantic-graph. It uses an iterative feedback loop where the planner proposes one question at a time, sees the answer, then decides whether to continue or synthesize.

---
title: Research graph
---
stateDiagram-v2
  [*] --> plan_next
  plan_next --> search_one: Has next question
  plan_next --> synthesize: Complete or max iterations
  search_one --> plan_next
  synthesize --> [*]

Key nodes:

• plan_next: Evaluates gathered evidence and either proposes the next question to investigate or marks research as complete
• search_one: Answers a single question using the knowledge base
• synthesize: Generates a final structured research report

Primary models:

• IterativePlanResult — planning decision (is_complete, next_question, reasoning)
• SearchAnswer — answer to a single question (query, answer, confidence, citations)
• ResearchReport — final report (title, executive summary, findings, conclusions, …)
• ConversationalAnswer — alternative output for chat integration (answer, citations, confidence)

Iterative flow:

• Each iteration: planner evaluates context → proposes one question → search answers it → loop back
• Planner can decompose complex questions (e.g., "benefits and drawbacks" → start with "benefits")
• Session context is used to resolve ambiguous references and inform planning
• Loop terminates when planner marks is_complete=True or max_iterations is reached

CLI Usage

# Basic usage
haiku-rag research "How does haiku.rag organize and query documents?"

# With document filter
haiku-rag research "What are the key findings?" --filter "uri LIKE '%report%'"

Python Usage

Basic example:

from haiku.rag.client import HaikuRAG
from haiku.rag.config import Config
from haiku.rag.agents.research.dependencies import ResearchContext
from haiku.rag.agents.research.graph import build_research_graph
from haiku.rag.agents.research.state import ResearchDeps, ResearchState

async with HaikuRAG(path_to_db) as client:
    graph = build_research_graph(config=Config)
    context = ResearchContext(original_question="What are the main features?")
    state = ResearchState.from_config(context=context, config=Config)
    deps = ResearchDeps(client=client)

    report = await graph.run(state=state, deps=deps)

    print(report.title)
    print(report.executive_summary)

With custom config:

from haiku.rag.client import HaikuRAG
from haiku.rag.config.models import AppConfig, ResearchConfig
from haiku.rag.agents.research.dependencies import ResearchContext
from haiku.rag.agents.research.graph import build_research_graph
from haiku.rag.agents.research.state import ResearchDeps, ResearchState

custom_config = AppConfig(
    research=ResearchConfig(
        provider="openai",
        model="gpt-4o-mini",
        max_iterations=5,
        max_concurrency=3,
    )
)

async with HaikuRAG(path_to_db) as client:
    graph = build_research_graph(config=custom_config)
    context = ResearchContext(original_question="What are the main features?")
    state = ResearchState.from_config(context=context, config=custom_config)
    deps = ResearchDeps(client=client)

    report = await graph.run(state=state, deps=deps)

Filtering Documents

Restrict searches to specific documents via the search_filter parameter:

# Set filter before running the graph
state = ResearchState.from_config(context=context, config=Config)
state.search_filter = "id IN ('doc-123', 'doc-456')"

report = await graph.run(state=state, deps=deps)

The filter applies to all search operations in the graph. See Filtering Search Results for available filter columns and syntax.