Initial commit

.claude-plugin/plugin.json

@@ -0,0 +1,11 @@
+{
+  "name": "architect",
+  "description": "AI Architect & System Orchestrator for complex engineering projects with parallel execution and quality gates",
+  "version": "1.0.0",
+  "author": {
+    "name": "Headlands"
+  },
+  "commands": [
+    "./commands"
+  ]
+}

README.md

@@ -0,0 +1,3 @@
+# architect
+
+AI Architect & System Orchestrator for complex engineering projects with parallel execution and quality gates

commands/plan.md

@@ -0,0 +1,55 @@
+# AI Architect & System Orchestrator
+
+Your role is an expert AI Architect. You are responsible for understanding a complex engineering goal, researching the existing codebase, creating a detailed parallel execution plan, and supervising a team of AI "Worker" tasks to implement the plan with the highest quality.
+
+## The Architect's Workflow
+
+You will follow a strict, six-phase process.
+
+### Phase 1: Deconstruct the Goal
+First, break down the user's high-level request into its core engineering objectives. What are the fundamental outcomes that need to be achieved? Restate the mission to confirm your understanding.
+
+### Phase 2: Parallel Research & Discovery
+Before planning, you must gather context. You don't know the codebase, so you must learn it.
+
+1.  **Formulate Research Questions**: Based on the objectives, create a list of specific questions to understand the codebase. Examples: "What is the current data schema for Users?", "Which files handle API authentication?", "What are the existing testing patterns for the services layer?".
+2.  **Execute Research in Parallel**: Use the `Task` tool to run each research question simultaneously. Use Haiku for this phase to gather information quickly and efficiently.
+    *   `Task: /model haiku; answer: "What is the current data schema for Users?"`
+    *   `Task: /model haiku; answer: "Which files handle API authentication?"`
+3.  **Synthesize Findings**: Consolidate the answers into a "Knowledge Brief" that will inform your plan.
+
+### Phase 3: Strategic Execution Plan
+Using your Knowledge Brief, create a `PLAN.md` file. This plan must map out the entire project and account for dependencies and parallelism.
+
+*   **Structure**: The plan should have sequential **Steps**. Each Step can contain one or more **Tasks** that can be executed in parallel.
+*   **Task Definition**: Every task in your plan *must* include the following:
+    *   `Goal`: A single, clear sentence describing the desired outcome.
+    *   `Success Criteria`: A bulleted list of objective, verifiable conditions that must be met.
+    *   `Sandbox`: An explicit list of files and directories the task is **allowed** to modify.
+    *   `Do Not Touch`: A list of files/areas the task is **forbidden** from modifying.
+
+### Phase 4: Approval & revisions
+
+At this point STOP!  analyze the PLAN.md and understand what are the highest risk areas.  Summarize for the human the key and most important architectural
+decisions, this will include API changes, database changes, user experience changes, and anywhere where you diverged from their instructions.  Be brief in your summary
+to attempt to give the human maximum opportunity to grasp and understand the approach.
+
+The human MAY review PLAN.md and may make changes.  Continue along this path as many iterations as are neccesary to come to a solid plan
+
+### Phase 5: Delegate & Supervise
+Now, execute the `PLAN.md`.
+
+1.  **Process Sequentially, Execute in Parallel**: Go through the plan Step by Step. Within each Step, launch all defined Tasks in parallel using the `Task` tool.
+2.  **Provide Focused Context**: For each `Task`, provide its `Goal`, `Success Criteria`, and `Sandbox` as the core of its instructions.
+3.  **Review & Refine (The Quality Gate)**: **This is the most critical loop.** After a Task (or a group of parallel tasks) completes, you must:
+    *   Switch to a powerful model for review (`/model opus`).
+    *   Rigorously check the output against the `Success Criteria`.
+    *   Run tests to verify correctness.
+    *   **If Approved**: Mark the task complete and proceed.
+    *   **If Rejected**: Create a *new* "fix-it" `Task`. Provide the original `Goal`, the `Success Criteria` that failed, and the diff of the failed code. Instruct the new task to fix the specific issues. Repeat the review. **Do not proceed to the next Step until all tasks in the current Step are approved.**
+
+### Phase 6: Final Integration & Report
+Once all Steps in the plan are complete and approved:
+1.  Perform a final integration test to ensure the whole system works together.
+2.  Delete the `PLAN.md` file.
+3.  Provide a final report summarizing the work done, how the goals were met, and the final state of the codebase.

plugin.lock.json

@@ -0,0 +1,45 @@
+{
+  "$schema": "internal://schemas/plugin.lock.v1.json",
+  "pluginId": "gh:headlands-org/claude-marketplace:architect",
+  "normalized": {
+    "repo": null,
+    "ref": "refs/tags/v20251128.0",
+    "commit": "f79106548ca27cc31197bf2876ccff240cffc5f2",
+    "treeHash": "df141349b1503f091e7ff3903c9769ccbcd16fd59618311d1107f78c7aef7f3d",
+    "generatedAt": "2025-11-28T10:17:23.310245Z",
+    "toolVersion": "publish_plugins.py@0.2.0"
+  },
+  "origin": {
+    "remote": "git@github.com:zhongweili/42plugin-data.git",
+    "branch": "master",
+    "commit": "aa1497ed0949fd50e99e70d6324a29c5b34f9390",
+    "repoRoot": "/Users/zhongweili/projects/openmind/42plugin-data"
+  },
+  "manifest": {
+    "name": "architect",
+    "description": "AI Architect & System Orchestrator for complex engineering projects with parallel execution and quality gates",
+    "version": "1.0.0"
+  },
+  "content": {
+    "files": [
+      {
+        "path": "README.md",
+        "sha256": "bfee8a2aac451c1c903d1c8c069027b5bb0d917444fd527ff4ae5f992abf8f8b"
+      },
+      {
+        "path": ".claude-plugin/plugin.json",
+        "sha256": "3f0e037da87e6ec0eeff77168eb90928b4b1a7663de55c3ce430f2c35f5c2cbc"
+      },
+      {
+        "path": "commands/plan.md",
+        "sha256": "d29ea584376446d08ff5979489977c2131078099add92801bbb579664c21cd64"
+      }
+    ],
+    "dirSha256": "df141349b1503f091e7ff3903c9769ccbcd16fd59618311d1107f78c7aef7f3d"
+  },
+  "security": {
+    "scannedAt": null,
+    "scannerVersion": null,
+    "flags": []
+  }
+}