lucianostraga

lucianostraga / sfagent-tools

Public

AI-powered testing for Salesforce Agentforce agents. Claude Code plugin.

10
1
100% credibility
Found Mar 31, 2026 at 11 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

A plugin for AI assistants that automates testing Salesforce Agentforce agents through simulated conversations, live transcripts, scored reports, and improvement recommendations.

How It Works

1
🔍 Discover the tool

While building your smart customer service agent in Salesforce, you learn about a helpful tool that lets an AI test it automatically without you writing any test scripts.

2
📥 Add it to your setup

You install the tool with one easy step, and it works with your existing Salesforce login and basic computer tools you already have.

3
🔗 Pick your test space

Choose your practice Salesforce area (like a sandbox) where your agent lives, keeping real customer areas safe.

4
💬 Chat with your AI helper

Simply tell your friendly AI, like Claude, 'test my agent' or share your business rules, and it starts exploring what your agent can do.

5
👀 Watch conversations live

Open a simple note file to see real-time chats between the AI tester and your agent, spotting issues as they happen.

📊 Review your results

Get a clear scorecard showing how well your agent handles talks, follows rules, and suggestions to make it even better.

Sign up to see the full architecture

4 more

Sign Up Free

Star Growth

See how this repo grew from 11 to 10 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is sfagent-tools?

SFAgent Tools is a TypeScript Claude Code plugin for AI-powered testing automation of Salesforce Agentforce agents. Tell Claude "Generate tests for my Agentforce agent," and it queries your org via sf CLI, designs scenarios from agent metadata, runs headless multi-turn conversations, and outputs scored reports with live transcripts and fix recommendations. Ditches manual YAML specs or Testing Center chats for minutes-long behavioral validation on routing, guardrails, and business rules.

Why is it gaining traction?

Unlike Salesforce Testing Center's rigid YAML or manual sessions, it uses Claude to auto-generate adaptive tests, probe edge cases like prompt injection, and score against custom expectations in sfagent-config.yaml. Zero-config install via `/plugin install sfagent-tools` leverages your existing sf auth, blocks prod orgs for safety, and spits out regression YAML—making ai-powered testing tools feel effortless for agentforce workflows.

Who should use this?

Salesforce devs crafting Agentforce service agents who need quick checks on topic routing and escalation before deploy. QA teams handling ai powered chatbot github projects, tired of missing multi-turn bugs in sandboxes.

Verdict

With 10 stars and 1.0% credibility score, it's immature but boasts thorough docs and open Apache 2.0 code—install in a dev org to test Agentforce agents without the YAML hassle. Solid for early adopters eyeing ai-powered testing frameworks.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.