Skip to content

getting-started.md

Latest commit

 

History

History
138 lines (96 loc) · 3.23 KB

getting-started.md

File metadata and controls

138 lines (96 loc) · 3.23 KB

AnyMind Getting Started

This guide focuses on running the code that is already in this repo.

Prerequisites

  • Python 3.10 to 3.12
  • Poetry
  • Credentials for the model provider you choose
  • Optional: Redis if you want the configured cache backend from config/model.json
  • Optional: ONNX build dependencies if you want local embedding assets

Install

poetry install

Optional ONNX assets:

poetry install --with onnx
python onnx_assets/build.py

Choose Configuration

By default, AnyMind looks for model config in these locations, in order:

  1. AM_MODEL_CONFIG
  2. ./model.json
  3. ./config/model.json
  4. repo-root/config/model.json
  5. ~/.config/anymind/model.json

MCP server config follows the same pattern, using AM_MCP_CONFIG and mcp_servers.json.

The repo includes example model configs:

Example environment setup:

export AM_MODEL_CONFIG=$PWD/config/model.openai.json
export AM_MCP_CONFIG=$PWD/config/mcp_servers.json

config/mcp_servers.json already includes the bundled local_tools MCP server, so no separate local-tools process is required.

Search tool credentials

If you want internet_search to work in the bundled local_tools MCP server, configure both upstream providers:

  • Kagi: set KAGI_API_KEY, or set search.kagi_api_key in the model config so the session factory exports it into the process environment
  • Scrapfly: set SCRAPFLY_API_KEY, or provide SCRAPFLY_API_KEY_SECRET_ARN

Example:

export KAGI_API_KEY=your_kagi_key
export SCRAPFLY_API_KEY=your_scrapfly_key

Important: without these values, the bundled local MCP server does not register internet_search. current_time and pdf_extract_text do not depend on those search credentials.

Run the CLI

Research-oriented workflow:

poetry run anymind --agent research_agent -q "Compare recent CPI trends across G7 nations"

Interactive CLI:

poetry run anymind --agent research_agent

SOP workflow from a file:

poetry run anymind --agent sop_agent -q "@/absolute/path/to/sop.json"

You can override model settings at runtime:

poetry run anymind \
  --agent research_agent \
  --provider openai \
  --model gpt-5.1 \
  --tools-policy planner \
  -q "Summarize the latest public SEC filing for Company X"

Run the API

Start the FastAPI service:

poetry run anymind serve --host 0.0.0.0 --port 8000

Important endpoints:

  • GET /health
  • POST /agents/run
  • POST /jobs
  • GET /jobs/{job_id}
  • POST /jobs/{job_id}/pause
  • POST /jobs/{job_id}/resume
  • POST /jobs/{job_id}/cancel

Example synchronous request:

curl -X POST http://127.0.0.1:8000/agents/run \
  -H 'content-type: application/json' \
  -d '{
    "message": "Compare recent CPI trends across G7 nations",
    "agent": "research_agent"
  }'

Tests

Run the test suite:

poetry run pytest -v

The repository already includes targeted tests for config loading, tool policy, evidence handling, usage tracking, caching, and MCP registry behavior.