repoburg 1.0.39 → 1.0.41

package/backend/.env

@@ -1,7 +1 @@
-# REPOBURG_PROJECT_PATH='/Users/celalertug/github/webcoder'
-# REPOBURG_PROJECT_PATH='/Users/celalertug/everest/local'
-# REPOBURG_PROJECT_PATH='/tmp/m/demo1'
-# REPOBURG_PROJECT_PATH='/Users/celalertug/github/typeorm'
- REPOBURG_PROJECT_PATH='/tmp/a/ecommerce/ecommerce'
-# REPOBURG_PROJECT_PATH='/tmp/a/fp'
-# REPOBURG_PROJECT_PATH='/Users/celalertug/everest/appdev3'
+REPOBURG_PROJECT_PATH='/Users/celalertug/WebstormProjects/hihiho'

package/backend/.repoburg/orchestrations/README.md

@@ -0,0 +1,68 @@
+# How to Create Orchestration Scripts
+
+Orchestrations automate multi-step tasks by executing a sequence of prompts. To create an orchestration, you must create two files in this directory: a script file (`.elf`) and a corresponding task file (`.md`).
+
+## 1. File Naming
+
+- Both files must share the same base name (e.g., `my-new-feature`).
+- The script file must have a `.elf` extension (e.g., `my-new-feature.elf`).
+- The task file must have a `.md` extension (e.g., `my-new-feature.md`).
+
+## 2. The `.elf` Script File
+
+This file defines the sequence of operations for the orchestration.
+
+### Commands
+
+Only two commands are supported:
+
+1.  **`set_initial_context('Template Name')`**
+    -   **Purpose:** (Optional, use sparingly) Overrides the session's default initial context template for the *first* `send_and_wait` command only. This is useful for highly specialized tasks that require a unique starting context.
+    -   **Usage:** Must be placed before the first `send_and_wait` command. In most cases, relying on the session's default templates is preferred for consistency.
+    -   **Example:** `set_initial_context('Default Initial Condensed')`
+
+2.  **`send_and_wait('Your prompt for this step')`**
+    -   **Purpose:** Sends a prompt to the AI and waits for the resulting actions to be automatically applied before proceeding to the next command. This is the core command for each step in the workflow.
+    -   **Usage:** The prompt string can be a single line using single quotes (`'`) or a multi-line string using backticks (``). Each `send_and_wait` represents one step.
+
+### Comments
+
+- Lines starting with `#` are treated as comments and are ignored.
+
+### Example `.elf` File (`sample-feature.elf`)
+
+```
+# This script creates a new React component and adds it to the app.
+# It uses a specific initial context, which is often not necessary.
+set_initial_context('Default Initial Condensed')
+
+send_and_wait(`Create a new React component file named Greeting.tsx in src/components. It should be a simple functional component that displays "Hello, World!".`)
+
+send_and_wait(`Now, import and render the new Greeting
+component in the main App component located at src/app/page.tsx.`)
+```
+
+## 3. The `.md` Task File
+
+This file provides a user-friendly checklist that mirrors the steps in the `.elf` script. The UI uses this file to track and display progress.
+
+### Format
+
+-   It is a GitHub Flavored Markdown checklist.
+-   Each `send_and_wait` command in the `.elf` file **must have a corresponding checklist item** in the `.md` file.
+-   The text of the checklist item must **exactly match** the prompt string within the corresponding `send_and_wait` command.
+-   Each task should start as unchecked: `[ ]`.
+
+### Example `.md` File (`sample-feature.md`)
+
+```markdown
+[ ] Create a new React component file named Greeting.tsx in src/components. It should be a simple functional component that displays "Hello, World!".
+[ ] Now, import and render the new Greeting component in the main App component located at src/app/page.tsx.
+```
+
+## Guidelines for Writing Prompts
+
+-   **Be Specific and Atomic:** Each `send_and_wait` prompt should describe a single, clear, and verifiable step.
+-   **Sequential Logic:** Write prompts in the order they should be executed. The orchestration runs them sequentially and depends on the success of the previous step.
+-   **Assume `auto_apply`:** Remember that all actions generated by the LLM will be applied automatically. Write prompts that are safe for direct execution.
+-   **Maintain Consistency:** The text in the `.md` checklist item must be identical to the prompt in the `.elf` file.