salesforce

An open, schema-driven language for configuring agent orchestration systems

40
10
100% credibility
Found Apr 16, 2026 at 40 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

AgentScript provides a declarative language and interactive tools for designing AI agent behaviors with visual editing, validation, and compilation.

How It Works

1
🔍 Discover AgentScript

You hear about a simple way to build smart AI helpers without complicated coding.

2
📥 Open the playground

Download and launch the friendly app to play with example agents right away.

3
✏️ Edit your agent

Type natural instructions in the easy editor with helpful hints and colorful checks.

4
🗺️ See the flow

Watch your agent's thinking path light up as an interactive map.

5
🧱 Add smart blocks

Drag and drop building pieces to make your helper smarter step by step.

6
Check everything

Spot any issues instantly and fix them with simple suggestions.

🚀 Your agent is ready

Share your custom AI helper that thinks and acts just how you want!

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 40 to 40 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 agentscript?

AgentScript is a schema-driven language for configuring agent orchestration systems, letting you define AI agents in a single file using blocks for state, execution flow, templating, and hooks. Developed by Salesforce for Agentforce but open for general use, it supports dialects like agentforce and agentfabric, with TypeScript libraries for parsing, linting, compiling, and LSP integration. Developers get a playground UI with Monaco editor, graph visualization, block builder, and AST explorer to test agent script recipes instantly.

Why is it gaining traction?

It decouples agent specs from runtimes, so scripts run unchanged on evolving systems—balancing determinism (explicit sets/if/transitions) with LLM autonomy. Full dev tools like VSCode extension, browser LSP, and real-time diagnostics make agentscript ai configs feel like editing TypeScript or YAML. Schema validation and dialect switching streamline configuring complex orchestration without boilerplate.

Who should use this?

Salesforce devs building Agentforce agents or Trailhead projects needing agentscript documentation and salesforce integration. Agent builders prototyping multi-topic flows with subagents, actions, and variables. TypeScript teams extending the toolchain for custom agent platforms.

Verdict

Promising open spec from Salesforce with solid agentscript documentation, playground, and LSP—install via pnpm, run `pnpm ui:dev` to try. At 40 stars and 1.0% credibility, it's early and runtime-tied to Salesforce, so evaluate for production carefully; great for experimentation or forking.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.