{"slug": "compile-for-coding-agents-a-isl-experimental", "title": "Compile for coding agents: A-ISL (experimental)", "summary": "A developer released A-ISL, an experimental open-source workflow that has a coding agent author a reviewable Spec in the classic ISL format (.isl.md) from a natural-language product description, which a human must accept before a separate skill compiles it into a target stack such as React, Blazor, WinForms or a C# library. The Spec never auto-chains into build output, preserving a human gate, and the same Spec can compile to multiple stacks without rewriting intent. In an internal closed-rules benchmark (tariffe-v2, N=5), Spec-to-compile matched a single prompt at 5/5 on React JS and 5/5 on C#, while an open DTO-style prompt dropped to 1/5 on first pass until a 'prefer records' rule was added; the developer frames this as a non-regression check, not a leaderboard, and says the expected payoff on large or multi-module projects is not yet demonstrated.", "body_md": "I got tired of two extremes: **one-shot prompts** that disappear into the chat, and **hand-written specs** nobody maintains.\n\n**A-ISL** is an experiment in the middle: you describe the product in natural language; a coding agent writes a **reviewable Spec** in classic [ISL](https://github.com/fra00/isl-specification) (`.isl.md`); you accept it; then you **explicitly** ask another skill to **compile** that Spec into `bin-<stack>/` (React, Blazor, WinForms, C# library, …).\n\nYou do **not** hand-author ISL or the app.\n\n`a-isl-spec-author``*.isl.md` only\n`a-isl-compile`` bin-<stack>/` via a small runtime (builder → queue → packet → lock)\nSpec never auto-chains into `bin/`. That’s the whole point: **human gate**, not autopilot.\n\nSame Spec can compile to more than one stack without rewriting intent — see the [sheet demo](https://github.com/fra00/a-isl-agents/tree/main/examples/sheet-workspace) (React / Blazor / WinForms).\n\nA-ISL does **not** invent the product. You decide behavior and accept the Spec. Wrong Spec that compiles is still “faithful.” That boundary is intentional.\n\nInternal bench (`tariffe-v2`, closed Domain/Logic rules, N=5):\n\n| Campaign | Stack | First full-pass | Final | \n|---|---|---|---|\n| A-ISL | react-js | 5/5 | 5/5 | \n| Single prompt | react-js | 5/5 | 5/5 | \n| A-ISL | csharp | 5/5 | 5/5 | \n| Prompt (open DTO style) | csharp | **1/5** | 5/5 | \n| Prompt + same “prefer records” rule | csharp | **5/5** | 5/5 | \n\n**Regression check, not a leaderboard.** On a small closed-rules task, Spec → Compile matches a single prompt (5/5 vs 5/5 on JS). On C#, a prompt that left DTO style open dropped to 1/5 on first pass; adding the same “prefer records” rule closed the gap. The expected payoff is on **large or multi-module projects** and on **change-via-Spec**. This bench does not demonstrate that; it only supports non-regression on a small closed oracle.\n\nFull honesty: [evaluation.md](https://github.com/fra00/a-isl-agents/blob/main/docs/evaluation.md).\n\nRuntime is agent-agnostic. Skills ship for Cursor, Claude Code, and GitHub Copilot (same text, synced). Day-to-day use and the bench were on **Cursor**; other hosts are not separately tested yet.\n\nModules, evolution via Spec, multi-stack. For a one-shot script, a normal prompt is enough — Spec→Compile costs more turns on purpose.\n\n**Prerequisites:** Node.js 18+; for Blazor / WinForms sheet demos, a .NET SDK (WinForms needs Windows targeting).\n\n```\ngit clone https://github.com/fra00/a-isl-agents.git\ncd a-isl-agents && npm install && npm --prefix runtime/lint install\n# open in your agent → /a-isl-spec-author → review → /a-isl-compile\n```\n\nTutorial: [docs/tutorial-first-project.md](https://github.com/fra00/a-isl-agents/blob/main/docs/tutorial-first-project.md)", "url": "https://wpnews.pro/news/compile-for-coding-agents-a-isl-experimental", "canonical_source": "https://dev.to/francesco_marconi_282058b/compile-for-coding-agents-a-isl-experimental-2djh", "published_at": "2026-11-02 12:21:16+00:00", "updated_at": "2026-11-02 12:23:42.004724+00:00", "lang": "en", "topics": ["ai-agents", "ai-tools", "developer-tools", "ai-products"], "entities": ["A-ISL", "ISL", "Cursor", "Claude Code", "GitHub Copilot", "React", "Blazor", "WinForms"], "also_reported_by": [], "alternates": {"html": "https://wpnews.pro/news/compile-for-coding-agents-a-isl-experimental", "markdown": "https://wpnews.pro/news/compile-for-coding-agents-a-isl-experimental.md", "text": "https://wpnews.pro/news/compile-for-coding-agents-a-isl-experimental.txt", "jsonld": "https://wpnews.pro/news/compile-for-coding-agents-a-isl-experimental.jsonld"}}