Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add stark-ai-de/agent-skills --skill architecture-zoomgit clone --depth 1 https://github.com/stark-ai-de/agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/stark-ai-de/agent-skills/architecture-zoom)<a href="https://agentmods.dev/skills/stark-ai-de/agent-skills/architecture-zoom"><img src="https://agentmods.dev/badge/skills/stark-ai-de/agent-skills/architecture-zoom/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/stark-ai-de/agent-skills/architecture-zoom"><img src="https://agentmods.dev/badge/skills/stark-ai-de/agent-skills/architecture-zoom.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00055 | $0.01544 |
| Opus 5.5 | $0.00022 | $0.00618 |
| Sonnet 5.5 | $0.00011 | $0.00309 |
| Haiku 4.5 | $0.00006 | $0.00154 |
Grade A, and why
architecture-zoom scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 6d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 84 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Architecture Zoom
Goal
Make the agreed whole product understandable, then describe only the detail needed for the next decision. Deliver a linked architecture blueprint and a path to complete, usable product increments, not an exhaustive specification or a tree of unfinished components.
This is one planning workflow. Architecture views describe structure; delivery views describe user outcomes. Neither implies extra processes, services, agents, or mandatory abstraction layers.
When to use
- Plan a new product, initial feature, or architecture before detailed implementation.
- Explain the whole system with optional drill-down into responsibilities and contracts.
- Cut a broad goal into releases that each fulfill a complete, bounded user promise.
- Update an existing blueprint after a material requirement or contract change.
When not to use
- A current implementation spec already resolves the requested change.
- A tiny correction needs no meaningful product or architecture decision.
- The request is an audit of existing code, direct refactoring, implementation, or formal proof.
- The user only wants brainstorming; do not turn that into mandatory specification work.
Inputs to inspect
Reuse the request, prior answers, product goals, release map, accepted local ADRs, relevant instructions, and existing work items. Inspect representative code and actual sources when needed to distinguish current state from intended behavior. Record material revisions and unavailable evidence; never invent paths, commands, interfaces, or test results.
Workflow
- Resolve scope and reuse. Identify the user promise, audiences, constraints, non-goals, delivery intent, and existing artifacts. Ask only unresolved material questions. Respect active Plan and observed permissions independently. Continue independent planning while a dependent decision is unresolved.
- Show the whole picture. Draft the overview using the blueprint template: complete agreed functional breadth, core journeys, system responsibilities, important cross-cutting requirements, and visible risks. Separate current state, target, assumptions, and proposals. Complete breadth does not mean predicting every future requirement.
- Refine contracts only where useful. Apply design contracts. Give modules stable IDs and one authoritative contract. Link parents, children, dependencies, and cross-cutting invariants; do not duplicate shared modules. Stop refining when the next decision has sufficient responsibilities, interfaces, assumptions, and acceptance criteria. Keep deferred detail and its revisit trigger visible.
- Cut complete product increments. Keep the architecture map separate from the delivery map. A release must fulfill its own user promise using existing capabilities plus its own work, never a future release. Prefer a narrow end-to-end path over completing database, backend, and UI separately. A ticket is not automatically an MVP. Bound necessary enabling work and POCs, link their consuming increment, and do not report them as usable products.
- Check bottom-up and across modules. Trace each material requirement through a journey, release, relevant contracts, and planned acceptance. Check compatibility, composition, error recovery, side effects, ownership, and global invariants. Surface circular reasoning and stronger-than-evidence claims. Prioritize missing core journey steps before polishing working ones; investigate existential feasibility risks early in bounded experiments.
- Review and stop. Classify findings as required corrections, material user trade-offs, or optional improvements. Use a bounded correction/recheck loop; default to one correction round and one targeted recheck, then surface unresolved blockers. Never weaken acceptance to call work complete. Optional improvements do not enlarge the current scope.
- Deliver or hand off. Present the overview, next complete user outcome, remaining decisions, and exact requested document scope. Reuse unchanged approvals; persist only authorized documents when the host permits, recheck state, and read back. Hand off the relevant contracts and work-item context, not a second whole-product specification. A planning-only request ends here.
What ships with it
7 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 6d ago First seen · 84 lines · 55 tokens per session scan A 69f231262445
architecture-zoom is a skill published in the GitHub repository stark-ai-de/agent-skills (8 stars, last pushed today), licensed Apache-2.0. It adds 55 tokens to every session and 1,544 once invoked, about $0.0002 per session on Opus 5.5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-10.
Other skills, from other repositories
warp-delegate
Delegate a coding task to the Warp Agent CLI (oz) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to hand implementation work to Warp - phrasings like "have Warp implement X", "delegate this to the Warp CLI", "run it through Warp", "use oz to…
claude-delegate
Delegate a coding task to a separate Claude Code CLI process or another Claude session as an implementer, then review its diff and land it yourself. Use only when the user explicitly asks to delegate implementation to Claude Code, another Claude session, or the claude CLI — for example, "have another Claude implement…
zcode-delegate
Delegate a coding task to the Z.AI ZCode CLI as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to hand implementation work to ZCode — phrasings like "have ZCode do X", "delegate this to ZCode", "run it through ZCode", or "use ZCode to implement/fix/refactor" — or…
cline-delegate
Delegate a coding task to the Cline coding agent CLI (cline) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to delegate implementation work to Cline - phrasings like "have Cline implement X", "delegate this to cline", "run it through Cline", or "use cline to…
kimi-delegate
Delegate a coding task to the Kimi Code CLI (kimi) as a background implementer, then review its diff and land it yourself. Use this whenever the user wants to hand implementation work to Kimi - phrasings like "have Kimi implement X", "delegate this to Kimi", "run it through Kimi Code", or "use Kimi to…
openlore-brainstorm
Transform a feature idea into an annotated story using a Domain Sketch or Constrained Option Tree. Use when asked to brainstorm, explore, or shape a feature before implementation.