CXAS SCRAPI CLI
CXAS SCRAPI is a Python package, cxas CLI and eight skills to build, lint and evaluate agents on Google CX Agent Studio.
- 116 ★ on the repository
- Last commit
- Apache-2.0
Extended access
Why this level
- `cxas push` changes an app in a Google Cloud project using the user's credentials.
- Some skills call billable Google Cloud services.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/cxas-scrapiDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- `cxas push` changes an app in a Google Cloud project using the user's credentials.
- Some skills call billable Google Cloud services.
This is third-party code. Review the repository files before installing.
What it does
CXAS SCRAPI is an open Google Cloud library for CX Agent Studio. The `Apps`, `Agents`, `Tools`, `Sessions` and other classes work with platform resources, `cxas lint` checks a project locally, and `cxas push` and `cxas pull` move an app between the platform and a folder. The `.agents/skills/` directory holds skills for Gemini CLI, Claude Code and Antigravity. `cxas-agent-foundry` runs the full cycle from requirements to a working agent: build, evaluations, debugging, iteration. Others cover narrow jobs: converting golden evaluations to simulations, migrating Dialogflow CX agents, analysing lost conversations from CCAI Insights, autolabelling rules, reports and dashboards.
Who it is for. Developers of conversational agents on Google Customer Engagement Suite who automate builds and tests through Gemini CLI or Claude Code.
Good fit when
- An agent must be built from product requirements together with its evaluations.
- A Dialogflow CX agent is being migrated to CX Agent Studio.
- A team analyses conversations the agent failed and looks for common causes.
- Before publishing, the app must pass `cxas lint` and run through simulations.
Not a fit when
- The platform is different: the library works only with CX Agent Studio.
- There is no access to a Google Cloud project.
- You need a ready-made cloud service: this is scripts and a CLI.
Example request
Build a support agent from these requirements, run `cxas lint` and prepare an evaluation set, but push nothing to the platform.Limitations
Needs Python 3.10 or newer and Google Cloud credentials: `gcloud auth application-default login` or a service account key with a Conversational Agents role. `cxas push` changes the app on the platform, so pass `cxas lint` first and work from a version snapshot with `--version-id`. Some skills depend on CCAI Insights and billable Google Cloud calls. The `cxas-agent-foundry` skill requires keeping a `todo.md` from its checklist and running the linter through a subagent.
How to disable. Delete the `cxas-*` folders from the client's skills directory and run `pip uninstall cxas-scrapi`. Revoke the service account key if you created one for the project.
FAQ
How do I install the library?
Run `pip install cxas-scrapi`. For repository development the README suggests `uv sync` and activating `.venv`.
Which skills are included?
Eight folders: `cxas-agent-foundry`, `cxas-autolabel-rules`, `cxas-composite-voice-agent-optimizer`, `cxas-configurable-dashboards`, `cxas-cuj-report-generator`, `cxas-dfcx-migration`, `cxas-loss-analysis` and `cxas-sim-eval`.
How do I authenticate?
In Colab, Cloud Functions and locally the library reads `gcloud` environment credentials. Or pass the path to a service account key as `creds_path`.
Can I export a version without touching the draft?
Yes. `cxas pull` accepts `--version-id` and exports an immutable snapshot instead of the live draft.
Is a separate Google Cloud setup needed?
Yes. You need a Google Cloud project with CX Agent Studio enabled and rights to work with apps; the library does not grant access.
Discussion
No questions or reviews yet.
Related
CLI toolsA library of 30-plus instruction skills for coding agents: planning, building, testing, hallucination guards and shipping
Tech Leads Club Agent Skills
Agent Skills
A vetted skill registry with a CLI that installs into twenty-plus agents and an MCP server for on-demand skill lookup
A CLI records coding agent sessions locally, replays what happened and shows what the trace could not prove.
The author's personal set: shared agent rules and dozens of skills for GitHub, the browser, macOS, messaging and releases, plus a sync script
Softaworks Agent Toolkit
Softaworks Agent Skills
A marketplace of skills, subagents and commands for coding agents: development, documentation, planning and workplace communication
A coding agent with a Python core and three front-ends: a script CLI, a terminal TUI and an Electron app, for OpenAI-style APIs
Comparison with similar tools
Cards that share the main search query, and how they differ.
| Tool | Official | Authentication | Repository stars | Main difference |
|---|---|---|---|---|
| CXAS SCRAPI (this card) | no | 116 | ||
| agentevals | no | not required | 169 | evaluates agent behaviour from OpenTelemetry traces on any platform; CXAS SCRAPI builds, lints and evaluates agents on CX Agent Studio specifically |
| agent-eval | yes | 280 | runs evaluation runs for coding agents; here evaluations are tied to goldens and simulations of the CX platform |
README in short
The library treats CX Agent Studio agents as a project of files that a linter can check and Git can store. The skills call the same `cxas` CLI.