architect
Sketch caller usage, types, signatures and module ownership before nontrivial implementation.
Skill by matthewyjiang
Install
Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.
npx agentmarket install skill/matthewyjiang-architect
Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.
Get the weekly diff
The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.
About
Reviews
Write a reviewNo reviews yet. Installed it? Reviews come from verified installers only. How reviews work
Permissions
What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.
- Tools
- No tools found
- Secrets
- None asked for
- Install scripts
- None found
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/architect
What's inside
- Skill name
architect- Description
- Sketch caller usage, types, signatures and module ownership before nontrivial implementation. Compare distinct designs, then implement only within the requested scope.
- License
- MIT
- Instructions
- 50 lines in SKILL.md · view on GitHub
Files (4)
- SKILL.md
- design-red-flags.md · references/
- rationale-template.md · references/
- runner-prompt.md · references/
SKILL.md (first lines)
# Architect Read [the workflow contract](../../references/workflow-contract.md). Design first; do not automatically write scaffolds or implement a design-only request. Track Ground → Sketch → Agree → Implement → Reassess with `todo` for substantial work. ## Ground Use `how` to understand touched subsystems. Use `why` when changing existing ownership or layering. Greenfield work still needs integration constraints. ## Sketch and compare Write caller-facing usage before deriving types, signatures and a module map. Explore structurally distinct viable shapes, not cosmetic variants of one idea. Compare at least two for a genuinely novel architectural choice; a narrow change with an established pattern needs no manufactured alternatives. Screen against `references/design-red-flags.md`: shallow modules, representation leaks, temporal decomposition, pass-through methods, split ownership, duplicate paths, importable internals and hand-synced lists. Prefer a smaller interface that hides real complexity, not a deeper call chain. Data structures and invariants come before logic. Subtract unnecessary state/layers before adding machinery. For substantial independent design perspectives, use separate calls to the existing `reviewer` per [review routing](../../references/review-routing.md). Same-model perspectives are normal; cross-model input is optional via per-call `model` or host delegation. Supply `references/runner-prompt.md`, `references/rationale-template.md`, the grounding and same task to each in one batch. They return sketches in their results, without editing files, creating worktrees or recursively running skills. Otherwise compare candidates in the parent. Do not claim model diversity for same-model or self-generated designs. Synthesize the best shape; record what won, what was borrowed/rejected and why. ## Agree and implement Respect requested checkpoints. Ask if a missing decision materially changes the result, compatibility or authority; otherwise proceed within the implementation request. A design-only task ends with the design. In-memory/textual `not implemented` bodies are fine for sketches, not an excuse to leave broken production commits. Fill in the chosen contract. Repeated new casts, special cases, shared-state workarounds or caller knowledge of internals suggest the sketch is wrong. Re-ground and redesign as though the discovered constraints were original requirements; do not keep bolting fixes on. Deliver verifiable increments under repository policy. Output caller usage, type/signature sketch, ownership/module map, rationale, alternatives/tradeoffs, risks, and next implementation step. Use the rationale template.
From matthewyjiang/rho-pstack, MIT.
More from matthewyjiang/rho-pstack
- 0.567score
how
matthewyjiang
Explain subsystem architecture, runtime flow, placement, ownership, or layering before changing code.
SkillAFreshness: AClaude Code +9 - 0.567score
interrogate
matthewyjiang
Optional adversarial cross-model review to challenge a design or find blind spots, using per-call agent models or host delegation.
SkillAFreshness: AClaude Code +9 - 0.561score
correct
matthewyjiang
Turn recurring agent mistakes into structural prevention: architecture, types, actionable lint, then behavior tests.
SkillAFreshness: AClaude Code +9 - 0.561score
blast-radius
matthewyjiang
Trace what a change could break beyond its diff and prove the load-bearing safety assumptions through real code.
SkillAFreshness: AClaude Code +9 - 0.536score
principle-boundary-discipline
matthewyjiang
Apply when wiring validation, error handling, or framework adapters.
SkillAFreshness: AClaude Code +9 - 0.536score
principle-build-the-lever
matthewyjiang
Apply to any non-trivial work, not just bulk work: edits, migrations, analyses, checks.
SkillAFreshness: AClaude Code +9 - 0.531score
principle-experience-first
matthewyjiang
Apply when product, UX, or feature-scope tradeoffs come up.
SkillAFreshness: AClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 17, 2026 (12 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/matthewyjiang/rho-pstack
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from matthewyjiang
- 0.567score
how
matthewyjiang
Explain subsystem architecture, runtime flow, placement, ownership, or layering before changing code.
SkillAFreshness: AClaude Code +9 - 0.567score
interrogate
matthewyjiang
Optional adversarial cross-model review to challenge a design or find blind spots, using per-call agent models or host delegation.
SkillAFreshness: AClaude Code +9 - 0.561score
correct
matthewyjiang
Turn recurring agent mistakes into structural prevention: architecture, types, actionable lint, then behavior tests.
SkillAFreshness: AClaude Code +9 - 0.561score
blast-radius
matthewyjiang
Trace what a change could break beyond its diff and prove the load-bearing safety assumptions through real code.
SkillAFreshness: AClaude Code +9 - 0.536score
principle-boundary-discipline
matthewyjiang
Apply when wiring validation, error handling, or framework adapters.
SkillAFreshness: AClaude Code +9 - 0.536score
principle-build-the-lever
matthewyjiang
Apply to any non-trivial work, not just bulk work: edits, migrations, analyses, checks.
SkillAFreshness: AClaude Code +9
Cite this page
architect skill. Agent Mart. https://agentmart.dev/skills/matthewyjiang-architect (last updated 2026-10-24)