@nomad-e/bluma-cli 0.0.61 → 0.0.62
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/main.js +10 -10
- package/package.json +1 -1
package/dist/main.js
CHANGED
|
@@ -1686,7 +1686,7 @@ When asked to perform tasks such as fixing bugs, adding features, refactoring, o
|
|
|
1686
1686
|
Objective: To independently implement and deliver a visually appealing, substantially complete, and functional prototype. Use all tools at your disposal to implement the application.
|
|
1687
1687
|
|
|
1688
1688
|
1. Understand the Requirements: Analyze the user request to identify key features, desired user experience (UX), visual aesthetics, application type/platform (web, mobile, desktop, CLI, library, 2D or 3D game), and explicit constraints. If critical information for initial planning is missing or ambiguous, ask concise and objective questions for clarification.
|
|
1689
|
-
2. Task Checklist: Create a series of well-structured tasks in the
|
|
1689
|
+
2. Task Checklist: Create a series of well-structured tasks in the to_do array, ensuring context and alignment with the project. Tasks should be designed considering: the type and main purpose of the application; and the main technologies to be used. The key features the application will offer and how users will interact with it. Design and UX approaches should prioritize beautiful, modern, and refined visual design, with special attention to user experience (UX)\u2014especially for UI-based applications.
|
|
1690
1690
|
- Ultimate Goal:
|
|
1691
1691
|
Ensure that each task contributes to a cohesive, functional, and visually appealing final product. For applications that require visuals (such as games or rich UIs), spend as much time as necessary planning and thinking through strategies for obtaining or generating placeholders (e.g., simple geometric shapes, procedurally generated patterns, or open-source resources, if feasible and licenses permit) to ensure a visually complete initial prototype. Ensure this information is presented in a structured and easy-to-understand format. - When the main technologies are not specified, give preference to the following:
|
|
1692
1692
|
- **Websites (Frontend):** NEXT.js (TypeScript) with Tailwindcss, incorporating Material Design or Shadcn principles for UI/UX.
|
|
@@ -1719,8 +1719,8 @@ Ensure that each task contributes to a cohesive, functional, and visually appeal
|
|
|
1719
1719
|
### IMPORTANT RULES
|
|
1720
1720
|
- Sending the initial message is mandatory and marks the turn start.
|
|
1721
1721
|
- Using the reasoning notebook is mandatory.
|
|
1722
|
-
- Breaking the task into **
|
|
1723
|
-
- Never include future steps in the **thought** field, only in the **
|
|
1722
|
+
- Breaking the task into **to_do** with the reflective problem-solving tool is mandatory.
|
|
1723
|
+
- Never include future steps in the **thought** field, only in the **to_do** checklist.
|
|
1724
1724
|
- Calling \`<agent_end_turn_rules>\` is mandatory to close the turn.
|
|
1725
1725
|
- Decline out-of-scope tasks professionally before calling \`<agent_end_turn_rules>\`.
|
|
1726
1726
|
- Process only one task per turn, never multiple concurrently.
|
|
@@ -1834,13 +1834,13 @@ CRITICAL: Your laptop (**reasoning_nootebook**) is your ORGANIZED MIND
|
|
|
1834
1834
|
---
|
|
1835
1835
|
|
|
1836
1836
|
<agent_end_turn_rules>
|
|
1837
|
-
This tool is mandatory, but MUST only be called when all tasks in \`
|
|
1837
|
+
This tool is mandatory, but MUST only be called when all tasks in \`to_do\` are fully completed.
|
|
1838
1838
|
|
|
1839
1839
|
Rules:
|
|
1840
1840
|
1. Never call this tool before all tasks are completed.
|
|
1841
|
-
2. It is strictly forbidden to call \`agent_end_turn\` if there are any pending tasks in \`
|
|
1841
|
+
2. It is strictly forbidden to call \`agent_end_turn\` if there are any pending tasks in \`to_do\`.
|
|
1842
1842
|
3. Before calling, always send a final message summarizing the completed work Turn.
|
|
1843
|
-
4. Verify that every task in the \`
|
|
1843
|
+
4. Verify that every task in the \`to_do\` array has a "completed" status before calling.
|
|
1844
1844
|
</agent_end_turn_rules>
|
|
1845
1845
|
|
|
1846
1846
|
|
|
@@ -2318,9 +2318,9 @@ CRITICAL: Your laptop (reasoning_nootebook) is your ORGANIZED MIND
|
|
|
2318
2318
|
- Jumping between unrelated subtasks
|
|
2319
2319
|
|
|
2320
2320
|
Important rule:
|
|
2321
|
-
Do not include future steps/to-dos in thought; put them strictly in
|
|
2321
|
+
Do not include future steps/to-dos in thought; put them strictly in to_do, using the mandated checklist markers.
|
|
2322
2322
|
|
|
2323
|
-
-
|
|
2323
|
+
- to_do: Checklist list of high-level upcoming tasks.
|
|
2324
2324
|
Format is mandatory:
|
|
2325
2325
|
- "\u{1F5F8}" \u2192 for tasks not yet done (pending)
|
|
2326
2326
|
- "[ ]" \u2192 for tasks already completed
|
|
@@ -2944,9 +2944,9 @@ var renderBlumaNotebook = ({
|
|
|
2944
2944
|
/* @__PURE__ */ jsx8(Text8, { color: "white", bold: true, children: "Reasoning:" }),
|
|
2945
2945
|
/* @__PURE__ */ jsx8(Box8, { marginLeft: 2, children: /* @__PURE__ */ jsx8(Text8, { color: "gray", children: thinkingData.thought }) })
|
|
2946
2946
|
] }),
|
|
2947
|
-
thinkingData.
|
|
2947
|
+
thinkingData.to_do && thinkingData.to_do.length > 0 && /* @__PURE__ */ jsxs8(Box8, { marginTop: 1, flexDirection: "column", children: [
|
|
2948
2948
|
/* @__PURE__ */ jsx8(Text8, { color: "white", bold: true, children: "Todos:" }),
|
|
2949
|
-
thinkingData.
|
|
2949
|
+
thinkingData.to_do.map((task, index) => /* @__PURE__ */ jsx8(Box8, { marginLeft: 2, children: /* @__PURE__ */ jsx8(Text8, { children: /* @__PURE__ */ jsx8(
|
|
2950
2950
|
Text8,
|
|
2951
2951
|
{
|
|
2952
2952
|
color: task.startsWith("\u{1F5F9}") ? "gray" : "white",
|