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 heliohq/ship --skill arch-designgit clone --depth 1 https://github.com/heliohq/shipWrote 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/heliohq/ship/arch-design)<a href="https://agentmods.dev/skills/heliohq/ship/arch-design"><img src="https://agentmods.dev/badge/skills/heliohq/ship/arch-design.svg" alt="Measured on agentmods" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00109 | $0.01806 |
| Opus 5 | $0.00055 | $0.00903 |
| Sonnet 5 | $0.00022 | $0.00361 |
| Haiku 4.5 | $0.00011 | $0.00181 |
Grade A, and why
arch-design 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 8d 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 — 167 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ship: Architecture Design
Think the design through before anything is written or built. The lenses below apply to any system: a service, a CLI, a frontend, a data pipeline, an agent runtime. Scale the depth to the decision — a single component with clear constraints needs the frame, a contract sketch, and trade-offs; a new system with unknowns needs every lens. Skipping a lens is a judgment call you record with its reason ("no trust boundary crossed — internal tool, single user"), never a silent omission.
Path note: ../shared/*.md references resolve against this skill's base
directory (announced as "Base directory for this skill" when the skill
loaded), not your working directory.
The quality bar for every lens: statements someone could prove wrong — named numbers, named failure behaviors, named rejected alternatives. Virtue words ("scalable", "robust", "flexible") with no test attached are filler.
Red Flag
Never:
- Jump to a solution before the frame (goals, non-goals, numbers) exists
- Present a design with zero numbers and zero rejected alternatives — that's a description, not a design
- Resolve a question from memory when the codebase can answer it
- Interrogate the user question-by-question — self-interview instead; user-owned calls become recorded assumptions
- Open many design branches at once — resolve dependencies in order
- Skip a lens silently — skip with a recorded reason, or don't skip
Method — Walk the Lenses as a Self-Interview
Interview yourself relentlessly about every aspect of the design until no unresolved question remains — the lenses below are the branches of the tree. Take questions one at a time, in dependency order: resolve a decision before opening the ones that build on it (storage before schema, contract before internals) — an answer stacked on an unresolved dependency is a guess, and opening many branches at once produces shallow parallel guesses. For each question, state your recommended answer, then resolve it with the strongest means available:
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.
- 8d ago First seen · 167 lines · 109 tokens per session scan A adb095d8007b
arch-design is a skill published in the GitHub repository heliohq/ship (92 stars, last pushed 2mo ago), licensed MIT. It adds 109 tokens to every session and 1,806 once invoked, about $0.0005 per session on Opus 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 2026-08-30.
Other skills, from other repositories
cesiumjs-core-utilities
CesiumJS core utilities and networking - Resource, Color, Event, Request, RequestScheduler, error handling, helper functions, feature detection. Use when fetching remote data, managing HTTP requests, working with colors, handling events, debugging errors, or using utility functions like defined, clone, or…
fec-data-fetching
A frontend guide for handling data that comes from a server, including typed requests, caching, refreshing, pagination, and updates. Server state means data owned by an API rather than by a page’s local controls.
fec-api-integration
A guide for connecting front-end code to back-end APIs, including typed clients, authentication, errors, uploads, and live connections such as SSE or WebSockets.
fec-react-project-standard
A guide for organizing medium-sized or large React and TypeScript projects. It defines boundaries between pages, business features, shared components, hooks, services, state, APIs, and utilities.
fec-source-driven-development
A development practice that bases frontend decisions on the current project, official documentation, source code, and tests. It is especially concerned with behavior that changes between versions.
fec-doc-sync
A workflow for keeping front-end project documentation aligned with the code and configuration that actually exist. It checks sources such as package files, routes, API clients, schemas, tests, continuous-integration settings, deployment files, and environment examples.