# andrewyng/translation-agent

Repository: https://github.com/andrewyng/translation-agent
Canonical: https://ross.abutalabs.com/products/translation-agent
Language: Python
License: MIT
License Family: permissive
Last push: 2024-08-04T06:14:56+00:00

## Health v2 (maintenance only)
Score: 24/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 35, longevity 60
- inputs: {"age_days": 843, "days_push": 759, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 5804, forks 708 (observed 2026-08-28T04:09:29.836493+00:00)

## What it is
A Python demonstration library by Andrew Ng implementing an agentic machine translation workflow: an LLM translates text, reflects on its own output, and applies suggestions to improve it. It is explicitly a research demo rather than production software.

## Use cases
- translate text with an llm using a reflection workflow
- control translation style and formality via prompts
- enforce consistent terminology with a glossary in machine translation
- translate for a specific regional dialect like latin american spanish
- generate parallel text corpora for training translation models

## When to choose
- you want steerable, prompt-controlled machine translation with custom style or terminology
- you are experimenting with agentic LLM workflows for translation research
- you need to generate training data (parallel corpora) for traditional MT systems

## When to avoid
- you need fast, cheap, production-grade translation at scale
- you require consistently better quality than commercial MT offerings
- you need a mature, well-supported translation product

## Facets
- artifact type: library
- maturity: experimental
- function: llm-inference, agent-framework, nlp, prompt-engineering
- domain: large-language-models
- platform: python, cross-platform
- tags: machine-translation, reflection-workflow, agentic-translation, llm, translation, natural-language-processing, ai-agents, localization

## Member repositories
- andrewyng/translation-agent (main) score 24

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:09:29.836493+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 2026-08-29T17:52:44.949944+00:00, confidence not recorded.
  - readme: https://github.com/andrewyng/translation-agent (fetched 2026-08-28T04:09:29.836493+00:00, sha 87fe70cc0ff1)
- Data as of 2026-08-30T08:39:29.467469+00:00.