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 rules/tencent-rtc/agent-skills/ui-modegit clone --depth 1 https://github.com/Tencent-RTC/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/rules/tencent-rtc/agent-skills/ui-mode)<a href="https://agentmods.dev/rules/tencent-rtc/agent-skills/ui-mode"><img src="https://agentmods.dev/badge/rules/tencent-rtc/agent-skills/ui-mode.svg" alt="Measured on agentmods" 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 | $0.01784 | $0.01784 |
| Opus 5 | $0.00892 | $0.00892 |
| Sonnet 5 | $0.00357 | $0.00357 |
| Haiku 4.5 | $0.00178 | $0.00178 |
Grade C, and why
ui-mode scanned grade C with 1 finding 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 3d 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.
Hidden instructionshighPrompt injection
Directives inside HTML comments, invisible characters or bidirectional overrides are read by the model and not by the person reviewing the file.
<!-- DO NOT EDIT — generated from ai-instructions/ by skills/trtc/room-builder/tools/render_ai_instructions.py. Edit the source markdown and re-run the renderer instead. --> How it starts
The opening of the file, as written. The whole thing — 159 lines — stays where its author put it; the contents beside it link to each section on GitHub.
When this rule applies
Only when .trtc-session.yaml at the repo root has ui_mode: official-roomkit
or ui_mode: full-ui. If the file is missing, or ui_mode is unset / null /
headless, this rule does not apply — fall back to whatever the tool's default
behavior is.
Medical new-project shortcut
When the user request is for medical online consultation (for example
1v1-video-consultation, remote diagnosis, online medical consultation,
doctor-patient video visit) and the task is to generate a brand-new
project rather than integrate into an existing app, bypass the normal full-ui
theme workflow entirely. Do not run trtc_prepare_ui.py, do not generate Vue
SFCs by hand, and do not run UI or medical verifiers.
Instead, directly copy the bundled medical template project:
skills/trtc/room-builder/templates/scenarios/medical-consultation/
to the user's local target project directory, preserving the project exactly
as packaged. In the handoff or integration notes, tell the customer to use
pnpm install for dependencies and pnpm dev for local development. Do not
recommend npm install / npm run dev; this medical template starts much
slower with npm and can show a blank first screen for a while.
This shortcut applies only to brand-new medical consultation projects. Existing project integration still follows the normal scenario / official-roomkit / full-ui rules below.
This is a terminal path. After choosing the bundled medical template, do not
enter the scenario/topic slice workflow, do not show a scenario capability table,
do not describe the project as a general-conference slice plan, and do not ask
about per-step auto-advance. Copy the template, mark the session completed, and
stop.
This is an internal execution rule, not user-facing copy. When explaining the
choice or reporting progress to the customer, do not mention "shortcut",
"bypass", "normal UI mode", "theme workflow", "CLAUDE.md", or any internal
rule file. Say that you are creating a complete 1v1 video consultation project
from the bundled medical template, and explain only the outcome and next
commands (pnpm install, pnpm dev).
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.
- 3d ago First seen · 159 lines · 1,784 tokens per session scan C e40545d5b6b3
ui-mode is a cursor rule published in the GitHub repository Tencent-RTC/agent-skills (12 stars, last pushed 13d ago), licensed MIT. It adds 1,784 tokens to every session, about $0.0089 per session on Opus 5. A static security scan graded it C with 1 finding (hidden instructions). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other cursor rules, from other repositories
commit
Conventional Commits specification and git commit workflow rules.
design
Design system, typography, color, accessibility (WCAG 2.2), responsive-first layout, and motion design standards.
dev-workflow
Monorepo developer workflow standards - mise tasks, git hooks, CI/CD, release automation.
frontend
React/Next.js frontend coding standards with shadcn/ui, Tailwind CSS v4, and FSD-lite architecture.
i18n-guide
Response language rules for agents and workflows.
infrastructure
Terraform infrastructure-as-code standards for multi-cloud (AWS, GCP, Azure, OCI) provisioning.