support for module .mjs files (#561)

* use tsx to run .mjs files

* mark esbuild as external

* add tsx dep

* added module test

* refine assertion

* packaging tsx

* alwasy importPrompt?

* add tidy examples

* updated globs

* various docs

* typo

* more docs tweak

* updatd samples

* notes about .mjs

* move a few tets to mjs

* updated test

* docs

* docs

* fix tests

* more mjs file

* added more docs

* package more deps

* add *.mjs to generated jsconfig

* calm down docs reviewer

* module voodoo

* updated jsconfig.json

* more logging of imports

* absolute path fix

* execution in server

* streaming of trace back to vscode

* handle cancellation

* add start stop script

* more tweaking of format

* more plumbing

* removing virtual file

* more fixes

* removing batch cli command, removing gpspecs

* more fixing

* typo

* updating fragment

* default files on phi3

* shorter summary

* tweak summary

* more plumbing

* better handling of building files

* catch exception in commands

* piping more messages to vscode

* wire cli info

* more logging on run

* fixing removal of spec

* more script fixes

* nire script fixes

* build fixes

* upgrading dependencies

* removing evalprompt

* test logging

* more logging

* Revert "removing evalprompt"

This reverts commit 6635661.

* revert eval import

* updated command.md

* use eval for .js files

* add genai-test command

* removing batch commands

* logging of csv output

* emoji for imports

* add ini test

.github/workflows/build-genai.yml

@@ -39,10 +39,7 @@ jobs:
             - name: download ollama docker
               run: docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
             - name: run test within scripts
-              run: yarn test:scripts --out-summary ./summary.txt --test-delay 10
-            - name: Append summary.txt to $GITHUB_STEP_SUMMARY
-              run: |
-                  cat packages/sample/summary.txt >> $GITHUB_STEP_SUMMARY
+              run: yarn test:scripts --out-summary $GITHUB_STEP_SUMMARY --test-delay 10
             - name: Add comment to PR
               uses: actions/github-script@v5
               if: github.event_name == 'pull_request'

.github/workflows/genai-commander.yml

@@ -62,6 +62,12 @@ jobs:
                   OPENAI_API_KEY: ${{ secrets.OPENAI_API_KEY }}
                   OPENAI_API_TYPE: ${{ secrets.OPENAI_API_TYPE }}
                   OPENAI_API_BASE: ${{ secrets.OPENAI_API_BASE }}
+            - name: start ollama
+              if: contains(github.event.comment.body, '/genai-test')
+              run: docker run -d -v ollama:/root/.ollama -p 11434:11434 --name ollama ollama/ollama
+            - name: run test within scripts
+              if: contains(github.event.comment.body, '/genai-test')
+              run: yarn test:scripts --out-summary $GITHUB_STEP_SUMMARY --test-delay 10
             - name: Archive genai results
               if: always()
               uses: actions/upload-artifact@v4

CONTRIBUTING.md

@@ -69,7 +69,7 @@ To compile and package the Visual Studio Code extension, run the `package` scrip
 yarn package
 ```
 
-You will find the built package files, `genaiscript.vsix` and `genaiscript.insiders.vsix`,
+You will find the built package files, `genaiscript.vsix`,
 in the `packages/vscode` folder.
 
 ## Release

demo/genaisrc/jsconfig.json

@@ -12,6 +12,7 @@
     },
     "include": [
         "*.js",
+        "*.mjs",
         "./genaiscript.d.ts"
     ]
 }

demo/web-app-basic-linux/summary_and_diagrams.md

@@ -12,7 +12,7 @@ stateDiagram
     AppServicePlan --> WebApp
 ```
 
-```js title="foo.genai.js"
+```js title="foo.genai.mjs"
 $`Write a one sentence poem.`
 ```
 

docs/genaisrc/jsconfig.json

@@ -12,6 +12,7 @@
     },
     "include": [
         "*.js",
+        "*.mjs",
         "./genaiscript.d.ts"
     ]
 }

docs/package.json

@@ -13,17 +13,17 @@
     "preview": "astro preview",
     "astro": "astro",
     "genai:test": "node ../packages/cli/built/genaiscript.cjs test src/**/*.md",
-    "genai:frontmatter": "node ../packages/cli/built/genaiscript.cjs batch frontmatter src/**/*.md --apply-edits",
-    "genai:technical": "node ../packages/cli/built/genaiscript.cjs batch technical src/**/*.md --apply-edits",
+    "genai:frontmatter": "for file in \"src/**/*.md\"; do\nnode ../packages/cli/built/genaiscript.cjs run frontmatter \"$file\" --apply-edits\ndone",
+    "genai:technical": "for file in \"src/**/*.md\"; do\nnode ../packages/cli/built/genaiscript.cjs run technical \"$file\" --apply-edits\ndone",
     "genai:alt-text": "node scripts/image-alt-text.mjs"
   },
   "dependencies": {
     "@astrojs/check": "^0.7.0",
-    "@astrojs/starlight": "^0.24.2",
-    "astro": "^4.10.3",
-    "typescript": "5.4.5"
+    "@astrojs/starlight": "^0.24.5",
+    "astro": "^4.11.3",
+    "typescript": "5.5.2"
   },
   "devDependencies": {
-    "zx": "^8.1.2"
+    "zx": "^8.1.3"
   }
 }

docs/src/content/docs/case-studies/image-alt-text.mdx

@@ -1,18 +1,19 @@
 ---
 title: Image Alt Text
 sidebar:
-  order: 10
+    order: 10
 description: Learn how to automatically generate descriptive alt text for images using OpenAI Vision model to enhance accessibility and SEO.
 keywords: alt text automation, image accessibility, SEO optimization, OpenAI Vision, Astro integration
 ---
-import { Image } from 'astro:assets';
-import { Code } from '@astrojs/starlight/components';
+
+import { Image } from "astro:assets"
+import { Code } from "@astrojs/starlight/components"
 import scriptSrc from "../../../../genaisrc/image-alt-text.genai.js?raw"
-import src from '../../../assets/debugger.png';
+import src from "../../../assets/debugger.png"
 import alt from "../../../assets/debugger.png.txt?raw"
 
-It is a best practice to provide an `alt` attribute for images. 
-This attribute is used to describe the image to users who are unable to see it. 
+It is a best practice to provide an `alt` attribute for images.
+This attribute is used to describe the image to users who are unable to see it.
 It is also used by search engines to understand the content of the image.
 
 ```html "alt"
@@ -30,7 +31,7 @@ However, this task can be tedious and developer are often tempted to skip it, or
 To solve this issue, we created a script that uses the OpenAI Vision model to analyze the documentation
 images and generate a description alt text.
 
-To start, we assume that the script is run on a single image file 
+To start, we assume that the script is run on a single image file
 and we use [defImage](/genaiscript/reference/scripts/images) to add it to the prompt context.
 
 ```js title="image-alt-text.genai.js"
@@ -55,18 +56,16 @@ The GenAIScript documentation uses Astro, which allows to author pages in [MDX](
 The code below shows how the generate alt text, stored in a separate text file, is injected in the final HTML.
 
 ```mdx
-
-import { Image } from 'astro:assets';
-import src from '../../../assets/debugger.png';
+import { Image } from "astro:assets"
+import src from "../../../assets/debugger.png"
 import alt from "../../../assets/debugger.png.txt?raw"
 
-<Image src={src} alt={alt}/>
-
+<Image src={src} alt={alt} />
 ```
 
 The `debugger.png` image shows the screenshot of a debugging session and the generate alt text file contents.
 
-<Image src={src} alt={alt}/>
+<Image src={src} alt={alt} />
 
 <Code code={alt} wrap={true} lang="txt" title="debugger.png.txt" />
 
@@ -76,7 +75,8 @@ Using the [batch](/genaiscript/reference/cli/batch) command, we can apply the sc
 in a single command.
 
 ```sh
-npx --yes genaiscript batch image-alt-text assets/**.png --apply-edits
+for file in assets/**.png; do
+  npx --yes genaiscript run image-alt-text "$file" --apply-edits
 ```
 
 To avoid regenerating the alt text, we also detect if a file exists in the script and cancel accordingly.

docs/src/content/docs/case-studies/seo-frontmatter.mdx

@@ -21,11 +21,11 @@ to merge the new front matter with the existing front matter.
 ## Batching over all files
 
 Once the script has been tuned on a few files, you can automate using
-the [CLI](/genaiscript/reference/cli). The CLI has a **batch** mode that applies a LLM generation
-on each file and a **--apply-edits** flag to apply the changes to the file.
+the [CLI](/genaiscript/reference/cli). The CLI has a **--apply-edits** flag to apply the changes to the file.
 
-```bash
-genaiscript batch frontmatter src/**/*.md --apply-edits
+```sh
+for file in src/**/*.md; do
+  npx --yes genaiscript run frontmatter "$file" --apply-edits
 ```
 
 You can run this command in your CI/CD pipeline to keep your SEO front matter up to date.
@@ -38,7 +38,7 @@ Add this command to your `package.json` to make it easier to run again.
 {
   ...
   "scripts": {
-    "genai:frontmatter": "genaiscript batch frontmatter src/**/*.md --apply-edits",
+    "genai:frontmatter": "for file in \"src/**/*.md\"; do\ngenaiscript run frontmatter \"$file\" --apply-edits\ndone",
   }
 }
 ```

docs/src/content/docs/faq.md

@@ -26,9 +26,6 @@ keywords: AI, script generation, performance, best practices, limitations
 -   **Can I use GenAIScript in IDEs other than VS Code?**
     Currently, GenAIScript is integrated with VS Code, but it can be written in any IDE. The VS Code extension, however, provides additional support for creating and debugging scripts.
 
--   **What is a GPSpec and how do I create one?**
-    A GPSpec is an optional markdown file that provides context for a GenAIScript. You can create one by writing a markdown document that lists arguments or includes more complex documentation.
-
 -   **What are foundation models and LLMs in the context of GenAIScript?**
     Foundation models and LLMs (Large Language Models) are AI models that GenAIScript can interact with to perform tasks like generating text or processing information.
 
@@ -60,9 +57,6 @@ keywords: AI, script generation, performance, best practices, limitations
 -   **How can I use GenAIScript to automate document translation?**
     One of our case studies illustrates the use of GenAIScript for translating document fragments between languages: [Translation Case Study](/genaiscript/case-studies/documentation-translations)
 
--   **Is it possible to generate executable code from natural language specifications using GenAIScript?**
-    Yes, you can use a GPSpec to define the task and a GenAIScript to translate the spec into code in a particular programming language.
-
 -   **Can I use GenAIScript to summarize documents or create dialogues from monologues?**
     Yes, LLMs are good at summarizing and can be used within GenAIScript to summarize documents or convert monologues into dialogues.
 

docs/src/content/docs/getting-started/automating-scripts.mdx

@@ -104,6 +104,7 @@ const { stdout: changes } = await host.exec("git", [
     "main",
     "--",
     ":!**/genaiscript.d.ts",
+    ":!**/jsconfig.json",
     ":!genaisrc/*",
     ":!.github/*",
     ":!.vscode/*",

docs/src/content/docs/getting-started/configuration.mdx

@@ -392,98 +392,6 @@ model running locally, while keeping the default `openai` model connection infor
 OLLAMA_PHI3_API_BASE=http://localhost:11434/v1
 ```
 
-## VSCode Copilot Insiders
-
-Visual Studio Code **Insiders** provides an
-**experimental support** to use your Copilot subscription
-to access a LLM inside of Visual Studio Code.
-
-:::caution[Limitations]
-
--   **This features requires a GitHub Copilot subscription.**
--   **This feature is still a [proposed api](https://github.com/microsoft/vscode/blob/main/src/vscode-dts/vscode.proposed.languageModels.d.ts) and requires the Insiders editor.**
--   The API changes daily and may not be stable.
--   Functions and Images are not supported.
--   The model configuration, temperature, max tokens, may not be available depending on the language model provider.
--   Some scripts may fail or behave differently due to internal alignment prompts from Copilot.
-
-:::
-
-<Steps>
-
-<ol>
-
-<li>
-
-Install [Visual Studio Code **Insiders**](https://code.visualstudio.com/insiders/).
-It can be installed side-by-side with the stable version of Visual Studio Code.
-
-</li>
-
-<li>
-
-Open your project in Visual Studio Code Insiders (green icon).
-
-<Image src={insidersSrc} alt={insidersAlt} />
-
-</li>
-
-<li>
-
-Download **genaiscript.insiders.vsix** from the [releases page](https://github.com/microsoft/genaiscript/releases/latest)
-to your project folder.
-
-<FileTree>
-
--   ...
--   **genaiscript.insiders.vsix**
-
-</FileTree>
-
-</li>
-
-<li>
-
-Right click on **genaiscript.insiders.vsix** and select **Install from VSIX**.
-
-</li>
-
-<li>
-
-Add a `.vscode-insiders/argv.json` file to enables loading the proposed APIs for the GenAIScript extension.
-
-```json title=".vscode-insiders/argv.json"
-{
-    "enable-proposed-api": ["genaiscript.genaiscript-vscode"]
-}
-```
-
-</li>
-
-<li>
-
-Restart Visual Studio Code Insiders to enable the extension.
-
-</li>
-
-<li>
-
-When you try to run a script, you will be prompted to authorize the Copilot subscription.
-After authorizing access, you will see a quick pick dialog to select one of the installed Language Model.
-
-<Image src={selectLLMSrc} alt={selectLLMAlt} />
-
-Note that this dialog is not used once you have a populated `.env` file.
-
-</li>
-
-</ol>
-
-</Steps>
-
-If you are using the Insiders build, make sure to update the **genaiscript.insiders.vsix** regularly
-by repeating the steps above. The daily builds of Insiders may contain breaking chagnes that require new builds
-of our extension.
 
 ## Next steps
 

docs/src/content/docs/getting-started/index.md

@@ -28,7 +28,7 @@ The following script generates a prompt that
 takes files (.txt, .pdf, .docx) as input and
 saves the summaries in another files.
 
-```js wrap title="summarize.genai.js" system=false assistant=true user=true
+```js wrap title="summarize.genai.mjs" system=false assistant=true user=true
 // context: define a "FILE" variable
 const file = def("FILE", env.files)
 // task: appends text to the prompt (file is the variable name)

docs/src/content/docs/getting-started/testing-scripts.mdx

@@ -17,7 +17,7 @@ to validate the output of the script.
 
 The tests are added as an array of objects in the `tests` key of the `script` function.
 
-```js title="proofreader.genai.js" wrap
+```js title="proofreader.genai.mjs" wrap
 scripts({
   ...,
   tests: {