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 runxhq/runx --skill taste-profilegit clone --depth 1 https://github.com/runxhq/runxWrote 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/runxhq/runx/taste-profile)<a href="https://agentmods.dev/skills/runxhq/runx/taste-profile"><img src="https://agentmods.dev/badge/skills/runxhq/runx/taste-profile/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/runxhq/runx/taste-profile"><img src="https://agentmods.dev/badge/skills/runxhq/runx/taste-profile.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00031 | $0.00911 |
| Opus 5 | $0.00015 | $0.00456 |
| Sonnet 5 | $0.00006 | $0.00182 |
| Haiku 4.5 | $0.00003 | $0.00091 |
Grade A, and why
taste-profile 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 5d 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 — 96 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Taste Profile
Turn demonstrated preferences into portable creative judgment. A taste profile captures what a person, team, product, or audience tends to choose, reject, and prioritize so a downstream agent can make better style decisions without pretending to know more than the evidence shows.
This is not a universal personality model. Taste is scoped to a surface and audience, changes over time, and may contain genuine tensions. The packet grants context only: it does not approve a design, publish content, purchase anything, or override accessibility and product constraints.
When to use it
Use taste-profile before design, writing, brand, product, or curation work
where repeated examples reveal a meaningful preference. It is useful when the
same operator wants several agents to share a stable aesthetic lens or when a
workflow must prove which preference context informed a result.
Do not use it to infer protected traits, diagnose a person, extrapolate from a
single weak example, or turn competitor work into permission to copy it. A
brand's communication rules belong in brand-voice; factual claims still need
their own evidence.
How it works
- Supply bounded evidence and label each item as a positive example, negative example, explicit preference, explicit dislike, or constraint.
- Runx normalizes the evidence, assigns stable local source references, and digests the complete admitted set through the native data boundary before synthesis. Evidence content is treated as data, so embedded instructions have no authority.
- The profile distinguishes strong repeated signals from tentative inferences and records tensions rather than forcing false consistency.
- Preferences become usable decision rules: favored qualities, disliked patterns, composition and density preferences, acceptable variation, and questions to ask when evidence does not decide.
- Deterministic finalization verifies that every claimed preference cites an admitted source reference and releases a packet bound to the native digest of the complete evidence set.
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.
- 5d ago First seen · 96 lines · 31 tokens per session scan A e26a157b9e92
taste-profile is a skill published in the GitHub repository runxhq/runx (84 stars, last pushed today), licensed Apache-2.0. It adds 31 tokens to every session and 911 once invoked, about $0.0002 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 2026-09-03.
Other skills, from other repositories
creative-claude-design
Design one-off HTML artifacts (landing, deck, prototype).
diagram-infographic
A guide for creating infographics, which are visual summaries of information, using 21 layout types and 21 visual styles.
desktop-brand-builder
Generate a branded Qwen Code desktop package from the Tauri desktop shell using a minimal brandId and logo. Use when the user wants a custom, white-label, or rebranded desktop client, installer, DMG/EXE/AppImage/deb, or one-click brand build on top of packages/desktop-shell.
ideate
Generate image-based alternatives, remixes, or new design directions from a Product Design brief. Use when the user asks for design variants, visual exploration, remixes, or image-generated approaches from provided context.
recording
Capture screen recordings and screenshots on any registered computer (macOS, Windows, Linux, HarmonyOS) and manage the recording library.
frontend-design
Produce intentional, responsive, accessible UI work and perform visual QA instead of generic component assembly.