- Home
- Skills
- Shotaiuchi
- Dotclaude
- Team Design
team-design_skill
- Shell
0
GitHub Stars
1
Bundled Files
2 months ago
Catalog Refreshed
4 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill shotaiuchi/dotclaude --skill team-design- SKILL.md6.1 KB
Overview
This skill composes and runs a multi-perspective design team to evaluate architecture, API, data, UX, infra, and security questions. It automatically detects the scope from input (PRs, issues, diffs, commits, paths, or free text), selects relevant panelists, runs parallel agent panelists, and synthesizes a decision document. The result is a structured trade-off analysis and a recommended approach that preserves dissenting views.
How this skill works
The skill first parses the provided arguments to detect the target scope and context (PR, issue, commit, diff, branch, file, or free text). It analyzes the topic to classify the design domain and then applies a selection matrix to choose which panelists to spawn. Selected panelists run in parallel, each returning a position; the lead agent synthesizes their debate into a final decision document with options, pros/cons, trade-offs, and rationale.
When to use it
- Evaluating architecture or design RFCs before major implementation
- Deciding between alternative system designs, APIs, or schemas
- Conducting structured pre-release reviews for PRs or branches
- Designing deployment, scaling, or compliance strategies
- Gathering multi-stakeholder perspectives for product or UX decisions
Best practices
- Provide precise context: include PR/issue numbers, commit range, or relevant files for accurate scope detection
- Define clear evaluation criteria (performance, cost, security, UX) up front
- Prefer including more context when uncertain — the selection rules err on thoroughness
- Limit team runs to a single focused question to keep outputs actionable
- Review the final decision document and record follow-up actions and owners
Example use cases
- Compare two service decomposition patterns for scaling and operational cost
- Decide API versioning and compatibility approach for a public SDK
- Select storage and migration strategy for a user-data model change
- Assess deployment topology, monitoring, and rollback strategy for a new feature
- Resolve UX flow trade-offs between accessibility and development effort
FAQ
It accepts PR or issue numbers, commit refs or ranges, unstaged/staged diffs, branch names, file or directory paths, or free-text descriptions.
Can I control which panelists are included?
Panelist selection follows explicit rules from the topic analysis, but you can influence inclusion by clarifying whether the topic is user-facing, infra-related, storage-heavy, or compliance-sensitive.