agentleFS
iflytekiflytek/astron-rpa/AGENTS.md

astron-rpa

AGENTS.md5.3k starsChanged 3 months agoApache-2.0View on GitHub

Guidance for AI coding agents working in the astron-rpa (AstronRPA) repository — an agent-ready RPA suite with out-of-the-box automation tools for desktop applications, web systems, and business-process automation. AstronRPA uses a server–client architecture: server components run via Docker, and the client (RPA engine + desktop app) is packaged and run on the machine that executes RPA tasks. The client toolchain is Windows-first. This is a multi-language monorepo driven by a unified Makefile (with modular includes under makefiles/). - backend/ —…

  • Reads credentials
  • Installs packages

We split this AGENTS.md into the rules it gives an agent

4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈1

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 3

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  • could be automatic: a setting could enforce it
  1. Very specific:

    Run make check and the relevant tests before opening a PR, and add or update tests for code you change.

    line 105
  2. Very specific:

    Create branches with the Make helpers (make new-feature name=<…>).

    line 120
  3. Very specific:

    Run make check and the relevant component tests before committing.

    line 121
  4. Distinctive:

    Keep PRs focused and write clear, conventional commit messages.

    line 123

≈ “Common” is an estimate: at least 1 in 5 of the 36,368 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

AGENTS.md

Guidance for AI coding agents working in the astron-rpa (AstronRPA) repository — an agent-ready RPA suite with out-of-the-box automation tools for desktop applications, web systems, and business-process automation.

AstronRPA uses a server–client architecture: server components run via Docker, and the client (RPA engine + desktop app) is packaged and run on the machine that executes RPA tasks. The client toolchain is Windows-first.

Project layout

This is a multi-language monorepo driven by a unified Makefile (with modular includes under makefiles/).

  • backend/ — server-side microservices (deployed via Docker Compose):
    • robot-service/, resource-service/, rpa-auth/ — Java / Spring Boot services (Maven; each ships checkstyle.xml, pmd-ruleset.xml, spotbugs-exclude.xml).
    • ai-service/, openapi-service/ — Python services (managed with uv).
  • engine/ — the Python RPA execution engine (managed with uv, linted with ruff): main.py, components/ (RPA components), servers/, shared/.
  • frontend/ — pnpm workspace monorepo (Vue 3 + TypeScript, Vitest, ESLint, i18n under locales/). Notable packages under frontend/packages/: electron-app (desktop client), web-app (web console), browser-plugin, auth-app, cli, components, shared.
  • docker/ — Docker Compose manifests for the server stack.
  • makefiles/ — modular Make components (common.mk, python.mk, java.mk, typescript.mk, go.mk, git.mk).
  • resources/ — packaged client resources. docs/, FAQ.md, BUILD_GUIDE.md — documentation.

Toolchain

  • Python 3.13.x + uv 0.8+ — RPA engine and Python services.
  • Node.js 22+ + pnpm 9+ — frontend workspace.
  • Java JDK 8+ + Maven — Java backend services.
  • Docker & Docker Compose — server deployment.
  • Windows client packaging also needs SWIG (Python↔C/C++ bindings, e.g. pywinhook) and 7-Zip (deployment archives). See BUILD_GUIDE.md.

Common commands

Quality and git workflow are driven by the root Makefile (run make help to see detected projects and targets):

  • make install-tools — install the formatting/linting toolchain.
  • make fmt — format across languages (fmt-python, fmt-java, fmt-ts, …).
  • make check (alias make lint) — format check + linters (check-python, check-java, check-ts, …). CI is check-only; it does not auto-fix.
  • make new-feature name=<short-name> / make new-bugfix name=<short-name> — branch helpers following the project's branch strategy.

Pre-commit hooks (install with pre-commit install) format the engine with ruff: uv run --project engine --dev ruff format ./engine.

Run the server (Docker)

cd docker
cp .env.example .env
# configure the public HTTPS names and copy the certificate/key to docker/certs/
# see docker/HTTPS_DEPLOYMENT.md for required variables
docker compose up -d mysql casdoor
# wait for Casdoor initialization, then synchronize this deployment's credentials
python3 scripts/sync-casdoor-credentials.py
docker compose up -d
docker compose ps

Gateway API: https://localhost · Casdoor auth: https://localhost:8443

Build the client (Windows)

# full build (engine + frontend desktop app) from the repo root
./build.bat --python-exe "C:\Program Files\Python313\python.exe"
./build.bat --skip-frontend   # engine only
./build.bat --skip-engine     # frontend only

The packaged installer is written to frontend/packages/electron-app/dist/.

Frontend (pnpm workspace, from frontend/)

pnpm install
pnpm dev:web         # run the web console in dev
pnpm dev:desktop     # run the desktop app in dev
pnpm build:desktop   # build the Windows desktop app

Testing instructions

There is no aggregate make test; run tests per component:

  • Frontend: pnpm test (Vitest) or pnpm test:run for a single run.
  • Python (engine, backend/ai-service, backend/openapi-service): uv run pytest within the project where tests exist (e.g. backend/ai-service has a tests/ directory).
  • Java (backend/robot-service, resource-service, rpa-auth): mvn test within the service directory.

Run make check and the relevant tests before opening a PR, and add or update tests for code you change.

Code style

Formatting and linting are wired through make fmt / make check:

LanguageFormat / Lint
Pythonruff format + ruff check (see engine/.ruff.toml)
JavaCheckstyle + PMD + SpotBugs (per-service configs)
TypeScript/VueESLint (frontend/eslint.config.mjs); pnpm lint / pnpm lint:fix

PR instructions

  • Create branches with the Make helpers (make new-feature name=<…>).
  • Run make check and the relevant component tests before committing; make sure pre-commit hooks pass.
  • Keep PRs focused and write clear, conventional commit messages.
  • For build and deployment details, see BUILD_GUIDE.md.

Download now (ZIP)

Trending

Less talk lately.about the whole repository

Dec 2025: 0 mentionsJan 2026: 0 mentionsFeb 2026: 0 mentionsMar 2026: 1 mentionApr 2026: 6 mentionsMay 2026: 2 mentionsJun 2026: 5 mentionsJul 2026: 6 mentionsAug 2026: 5 mentionsSep 2026: 5 mentionsOct 2026: 4 mentionsNov 2026: 0 mentions
Dec 2025Jun 2026Nov 2026
Mentions a month of the iflytek/astron-rpa repository that our weekly crawl found. There is no adoption data for this file yet.

Who's talking about iflytek/astron-rpa

The biggest 5 of 36 mentions linking to the repository that our crawl found. Each line is their own headline.

See all 36

Talked about on GitHub (36 times).

On GitHub · 36

31 more on GitHub
  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 28,693 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.