5,501,259 AI resources indexedWorking toward the world’s largest AI resource index.

PDF

architecture - Architecture Brief Methodology

Internal methodology for the architecture discovery stage. Produces a concise Architecture Brief for product and planning handoff.

Installable skillApache-2.0By GobbyAIUpdated Oct 28, 2026
INTENDED FOROpenClawInstall in a listed agent.

instructions.md

---
name: architecture-d595657f247b
description: Internal methodology for the architecture discovery stage. Produces a concise Architecture Brief for product and planning handoff.
version: "1.0.0"
category: methodology
internal: true
triggers: architecture brief, technical design, interfaces
metadata:
  gobby:
    audience: agent
    depth: 0
---

# architecture - Architecture Brief Methodology

Use this skill when acting as the `architect` discovery-stage agent.

## Purpose

Translate ideation and research into technical direction. Capture the drivers,
decisions, components, interfaces, and trade-offs that planning must preserve.

## Method

1. Read the task and prior discovery marker blocks.
2. Identify architectural drivers: data shape, lifecycle, integration points,
   safety, performance, compatibility, and migration constraints.
3. Make decisions only where evidence supports them.
4. Describe the smallest useful component/interface boundaries.
5. Surface trade-offs and unresolved questions that planning must account for.

## Output

Write only this section shape:

```markdown
## Architecture Brief

### Drivers
- <technical driver>

### Decisions
- <decision and reason>

### Components
- <component or module responsibility>

### Interfaces
- <API, storage, MCP, CLI, or file contract>

### Trade-offs
- <trade-off>

### Open Questions
- <question or "None">
```

Do not write implementation tasks or PRD acceptance criteria. Keep the brief
actionable for the planner.

Discuss this resource

Platform guidance

Imported from the listed AgentSkill.sh category. Complete supplied package files and notices are retained. Required external tools and credentials are not installed. No bundled code was executed during import. Individual agent task behavior has not been tested.