# facebookresearch/diplomacy_cicero

Code for Cicero, an AI agent that plays the game of Diplomacy with open-domain natural language negotiation.

Repository: https://github.com/facebookresearch/diplomacy_cicero
Canonical: https://ross.abutalabs.com/products/diplomacy_cicero
Language: Python
License: NOASSERTION
License Family: other
Archived: true
Last push: 2025-04-17T19:14:56+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 17, release rhythm 35, longevity 100
- inputs: {"age_days": 1400, "days_push": 503, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases, archived, no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1428, forks 168 (observed 2026-08-28T04:04:41.868822+00:00)

## What it is
Research code and model checkpoints for Cicero and Diplodocus, AI agents that play the board game Diplomacy at human level by combining language models with strategic planning and reinforcement learning. Built on ParlAI with Python training code and C++ components for search.

## Use cases
- reproduce the Cicero Diplomacy AI agent from the Science paper
- train reinforcement learning agents for no-press Diplomacy
- study language model negotiation dialogue in games
- run self-play RL training for strategic board games
- benchmark strategic planning agents against human-level play
- fine-tune language models for game negotiation

## When to choose
- you want to reproduce or extend published Diplomacy AI research
- you need checkpoints and training code for language-model-based game agents
- you're researching combining LLMs with planning and RL

## When to avoid
- you want a ready-to-play Diplomacy game client
- you need a general-purpose game AI framework
- you lack GPU resources or ML research experience

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, reinforcement-learning, nlp, llm-training, agent-framework, game-engine, simulation
- domain: artificial-intelligence, reinforcement-learning
- platform: python
- tags: game-ai, diplomacy, language-models, strategic-reasoning, research-code, facebook-research, negotiation, natural-language-processing, game-development, research, linux, gpu

## Member repositories
- facebookresearch/diplomacy_cicero (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:41.868822+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:37:19.072918+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/diplomacy_cicero (fetched 2026-08-28T04:04:41.868822+00:00, sha 8a6345444c41)
- Data as of 2026-08-30T08:39:29.467469+00:00.