riktar

riktar / slang

Public

A declarative meta-language for orchestrating multi-agent workflows. Readable by humans. Executable by LLMs. Portable across models.

riktar.github.ioslang agentic-ai agentic-ai-development agentic-engineering agentic-workflow llm-agent
15
0
100% credibility
Found Mar 18, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
TypeScript
AI Summary

SLANG is a plain-text language for describing multi-agent AI workflows that runs in any AI chat or via optional desktop tools.

How It Works

1
💡 Discover SLANG

You stumble upon SLANG, a friendly way for anyone on your team to describe AI agent teamwork without needing to code.

2
🧠 Paste into AI chat

Grab a ready prompt, add your simple workflow description, and drop it into ChatGPT or Claude – agents start chatting and working right away!

3
✏️ Describe your agents

Write what each agent does in easy steps like 'researcher gathers info then sends to analyst' – your whole team can read and edit it together.

4
Pick your playground
⚡
Quick chat test

Keep pasting and tweaking in your AI chat for instant results.

🔧
Full setup

Click to launch a web playground or desktop tool for editing and running.

5
👥 Team edits together

Share the file – PMs tweak logic, analysts add details, everyone sees the flow clearly.

6
🎮 Watch it run live

See agents connect, pass info, and finish tasks in a visual playground with real results.

✅ AI workflows owned by all

Your team now runs custom AI automations reliably, no more waiting on developers.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 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 slang?

Slang is a declarative meta-language for orchestrating multi-agent LLM workflows, readable by teams and executable by any model via riktar/slang on GitHub. Define flows with agents using stake (produce/send), await (receive), and commit primitives—no Python or TypeScript boilerplate. Use the CLI (`slang run flow.slang`), web playground for visualization, or zero-setup mode by pasting into ChatGPT; supports tools, parallelism, and structured outputs.

Why is it gaining traction?

Unlike SDKs like LangChain or CrewAI that lock teams into code, slang lets PMs edit declarative pipelines directly, with LLMs generating them like text-to-SQL. Static analysis catches deadlocks, LSP adds IDE completions/hovers, and the playground graphs dependencies instantly. Portable across 300+ OpenRouter models, with checkpointing for production.

Who should use this?

PMs and analysts building AI research or review pipelines without devs. Devs prototyping multi-agent systems, like researcher-analyst-critic flows with web_search tools. Teams needing shared workflow docs that double as executable slang github llm specs.

Verdict

Promising for declarative pipelines github but early-stage (15 stars, 1.0% credibility)—docs and playground shine, tests cover core paths, yet lacks battle-tested scale. Try the playground for quick wins; skip for mission-critical unless you contribute.

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.