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 agentmods add commands/f0d010c/stark/stark-referencegit clone --depth 1 https://github.com/f0d010c/starkWhat 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 | $0.00021 | $0.00255 |
| Opus 5 | $0.00010 | $0.00128 |
| Sonnet 5 | $0.00004 | $0.00051 |
| Haiku 4.5 | $0.00002 | $0.00026 |
Grade A, and why
stark-reference 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 2d 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.
What it actually says
/stark-reference
Legacy Claude Code command. In Codex, ask:
Use stark reference analysis for this flow before designing.
Read:
references/ui-patterns/reference-analysis.mdreferences/ui-patterns/asset-selection.md- the relevant UX or platform skill
Inspect 3-5 shipped references when available. Use official docs, live public pages, first-party screenshots, or approved reference libraries. If using Mobbin/Figma/reference screenshots, extract structure only.
Output:
Reference extraction brief
- Pattern chosen: ...
- Why it fits this user's job: ...
- Structural decisions to borrow: ...
- Decisions to reject: ...
- State/responsive requirements: ...
- Asset/source constraints: ...
Rules:
- Do not pixel-clone.
- Do not copy brand, artwork, screenshots, copy, or proprietary layouts.
- Name what you learned from each reference in product terms: job, IA, hierarchy, state, interaction, recovery, responsive behavior.
- Use the brief as input to UX/UI implementation.
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.
- 2d ago First seen · 39 lines · 21 tokens per session scan A c01ca098f1ce
stark-reference is a command published in the GitHub repository f0d010c/stark (25 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 21 tokens to every session and 255 once invoked, about $0.0001 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 commands, from other repositories
od-contribute
Open a first-contribution PR (or bug issue) on nexu-io/open-design — works for non-coders too.
OPSX: Explore
Enter explore mode - think through ideas, investigate problems, clarify requirements.
/opsx-propose
Propose a new change - create it and generate all artifacts in one step.
anti-slop-designer
Use when a UI looks machine-made rather than decided. Violet gradients, glassmorphism everywhere, identical cards in a grid, untouched shadcn or Material defaults, emoji bullets, or copy stuffed with seamless and unlock.
design-enforce
Use when code has drifted off the design system. Raw hex values and pixel numbers written where a token already exists, one-off components, library defaults left untouched. Finds every bypass and fixes it.
design-grill
Use when a design keeps getting reinvented every session. Interviews you until you share one understanding, then writes the vocabulary and the binding decisions into the repo so they survive the next context window.