Markdown Reader

Pass MarkdownReader to Knowledge.insert() to load a Markdown file.

markdown_reader.py

from pathlib import Path

from agno.agent import Agent
from agno.knowledge.knowledge import Knowledge
from agno.knowledge.reader.markdown_reader import MarkdownReader
from agno.models.openai import OpenAIResponses
from agno.vectordb.pgvector import PgVector

db_url = "postgresql+psycopg://ai:ai@localhost:5532/ai"

knowledge = Knowledge(
    vector_db=PgVector(
        table_name="markdown_documents",
        db_url=db_url,
    ),
)

agent = Agent(
    model=OpenAIResponses(id="gpt-5.2"),
    knowledge=knowledge,
    search_knowledge=True,
)

if __name__ == "__main__":
    markdown_path = Path("tmp/project.md")
    markdown_path.parent.mkdir(parents=True, exist_ok=True)
    markdown_path.write_text(
        "# Project Atlas\n\nProject Atlas tracks deployment readiness.\n",
        encoding="utf-8",
    )
    knowledge.insert(path=markdown_path, reader=MarkdownReader())
    agent.print_response("What does Project Atlas track?", markdown=True)

Run the Agent

Set up your virtual environment

uv venv --python 3.12
source .venv/bin/activate

Install dependencies

uv pip install -U "agno[markdown]" openai pgvector psycopg sqlalchemy

Export the API key

export OPENAI_API_KEY=your_openai_api_key_here

Run PgVector

docker run -d \
  -e POSTGRES_DB=ai \
  -e POSTGRES_USER=ai \
  -e POSTGRES_PASSWORD=ai \
  -e PGDATA=/var/lib/postgresql \
  -v pgvolume:/var/lib/postgresql \
  -p 5532:5432 \
  --name pgvector \
  agnohq/pgvector:18

Run the agent

python markdown_reader.py

Reader Parameters

Parameter	Type	Default	Description
`chunking_strategy`	`Optional[ChunkingStrategy]`	`MarkdownChunking()` when available	Strategy used to chunk the Markdown content

MarkdownReader also accepts the base Reader constructor parameters.

The agno[markdown] extra installs aiofiles. The reader selects MarkdownChunking only when its optional dependencies are available; otherwise it uses FixedSizeChunking. To enable Markdown chunking, also install unstructured and markdown as described in Markdown Chunking.

MarkdownReader.async_read() uses aiofiles when it is installed and falls back to synchronous file I/O otherwise.

Next Steps

Task	Guide
Configure Markdown chunking	Markdown Chunking
Choose another chunking strategy	Chunking Overview
Compare available readers	Readers Overview