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 Tencent-RTC/agent-skills --skill trtc-voice-ai-startergit 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/skills/tencent-rtc/agent-skills/trtc-voice-ai-starter)<a href="https://agentmods.dev/skills/tencent-rtc/agent-skills/trtc-voice-ai-starter"><img src="https://agentmods.dev/badge/skills/tencent-rtc/agent-skills/trtc-voice-ai-starter/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/tencent-rtc/agent-skills/trtc-voice-ai-starter"><img src="https://agentmods.dev/badge/skills/tencent-rtc/agent-skills/trtc-voice-ai-starter.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 | $0.00207 | $0.07539 |
| Opus 5.5 | $0.00083 | $0.03016 |
| Sonnet 5.5 | $0.00041 | $0.01508 |
| Haiku 4.5 | $0.00021 | $0.00754 |
Grade B, and why
trtc-voice-ai-starter scanned grade B with 2 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 4d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Asks for rootmediumPrivilege escalation
A mod that escalates privileges can change anything on the machine, not only the project.
cd "$TEMPLATE_ROOT" && (test -f .env || cp .env.example .env) && chmod 600 .env Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
sleep 6 && curl -sS http://localhost:3000/health How it starts
The opening of the file, as written. The whole thing — 569 lines — stays where its author put it; the contents beside it link to each section on GitHub.
TRTC Voice AI Starter Kit
§0 AGENTS.md Priority
If the user's CWD has a template/-generated AGENTS.md:
- Architecture / convention questions → trust
AGENTS.md - Skill workflow / setup questions → trust this file
- Safety red lines → both must agree; mismatch = bug, warn user
§1 Entry Guards
Check before §5. If any row matches, stop and redirect immediately.
| Condition | Action |
|---|---|
| User mentioned iOS as target | Tell user: "Guided integration is currently available for Web only. For iOS, you can follow the official documentation to implement it yourself: https://trtc.io/document/69000.md" Stop. |
| User mentioned Android as target | Tell user: "Guided integration is currently available for Web only. For Android, you can follow the official documentation to implement it yourself: https://trtc.io/document/69001.md" Stop. |
§2 Interaction Language Detection
Detect once from the triggering message. Store as interaction_lang (zh or en).
All AI guidance text, options, and summaries follow interaction_lang. Update if user switches.
§3 Prompt Reporting
Host bootstrap handles reporting. On TRTC_REPORTING_NOTICE_REQUIRED_V1: finish answer first, host hook displays runtime/continuation-notice.md. Fallback: stdin prompt once if host did not record.
§4 Path Baseline
| Variable | Meaning |
|---|---|
SKILL_ROOT |
This skill's directory (contains SKILL.md) |
TEMPLATE_ROOT |
$SKILL_ROOT/template |
PROJECT_ROOT |
User's workspace root |
Fallback if SKILL_ROOT unknown:
find "$PWD" -maxdepth 5 -name SKILL.md -path '*voice-ai-starter*' 2>/dev/null | head -1
Always substitute path variables with absolute paths. Never ask user to move directories.
§5 Scene Identification and Routing
First action after language detection — run before any environment checks. Present a single-choice question in interaction_lang:
If interaction_lang == zh:
"你好！在开始之前，先告诉我你想搭建哪种 AI 语音对话体验，我来帮你选择最合适的方式。"
What ships with it
60 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.
- README.md 8.6 KB
- README.zh-CN.md 7.2 KB
- references/extensions.md 5.3 KB
- references/provider-config.md 6.1 KB
- template/.env.example 7.2 KB
- template/.gitignore 255 B
- template/AGENTS.md 6.0 KB
- template/package.json 738 B
- template/pnpm-lock.yaml 84 KB
- template/pnpm-workspace.yaml 29 B
- template/README.md 3.9 KB
- template/scripts/switch-provider.mjs 7.7 KB runs code
- template/scripts/verify-credentials.mjs 5.5 KB runs code
- template/server/package.json 651 B
- template/server/src/config/overrides.test.ts 2.6 KB runs code
- template/server/src/config/overrides.ts 2.7 KB runs code
- template/server/src/config/providers.ts 6.5 KB runs code
- template/server/src/config/voice.test.ts 4.6 KB runs code
- template/server/src/config/voice.ts 4.1 KB runs code
- template/server/src/data/.gitkeep 0 B
- template/server/src/env.ts 4.8 KB runs code
- template/server/src/index.ts 3.1 KB runs code
- template/server/src/logger.ts 1.1 KB runs code
- template/server/src/routes/agent.ts 5.5 KB runs code
- template/server/src/routes/schemas.test.ts 2.8 KB runs code
- template/server/src/routes/schemas.ts 3.6 KB runs code
- template/server/src/routes/transcription.ts 3.4 KB runs code
- template/server/src/routes/userSig.ts 3.2 KB runs code
- template/server/src/routes/webhook.ts 6.6 KB runs code
- template/server/src/tencent/callbackEvents.ts 2.5 KB runs code
- template/server/src/tencent/client.ts 2.3 KB runs code
- template/server/src/tencent/endpoint.test.ts 1.3 KB runs code
- template/server/src/tencent/endpoint.ts 2.1 KB runs code
- template/server/src/tencent/types.ts 1.4 KB runs code
- template/server/src/tencent/userSig.ts 1.5 KB runs code
- template/server/src/tencent/webhook.test.ts 3.6 KB runs code
- template/server/src/tencent/webhook.ts 3.7 KB runs code
- template/server/tsconfig.json 169 B
- template/server/tsconfig.test.json 232 B
- template/start.sh 2.6 KB runs code
- template/tsconfig.base.json 298 B
- template/web/index.html 300 B
- template/web/package.json 577 B
- template/web/src/App.css 6.4 KB
- template/web/src/App.tsx 3.1 KB
- template/web/src/components/ControlBar.tsx 1.1 KB
- template/web/src/components/LEDIndicator.tsx 543 B
- template/web/src/components/StatusPanel.tsx 2.5 KB
- template/web/src/components/SubtitleStream.tsx 1.2 KB
- template/web/src/hooks/useAgent.ts 1.6 KB runs code
- template/web/src/hooks/useCustomMessage.ts 1.8 KB runs code
- template/web/src/hooks/useTRTC.ts 4.9 KB runs code
- template/web/src/lib/api.ts 3.5 KB runs code
- template/web/src/lib/conversation.test.ts 10 KB runs code
- template/web/src/lib/conversation.ts 5.0 KB runs code
- template/web/src/lib/session.test.ts 4.0 KB runs code
- template/web/src/lib/session.ts 2.2 KB runs code
- template/web/src/main.tsx 284 B
- template/web/src/styles/styles.test.ts 2.3 KB runs code
- template/web/src/styles/tokens.css 2.1 KB
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.
- 4d ago First seen · 569 lines · 207 tokens per session scan B 014de2abb2aa
trtc-voice-ai-starter is a skill published in the GitHub repository Tencent-RTC/agent-skills (15 stars, last pushed 4d ago), licensed MIT. It adds 207 tokens to every session and 7,539 once invoked, about $0.0008 per session on Opus 5.5. A static security scan graded it B with 2 findings (asks for root, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-24.
Other skills, from other repositories
cometchat-ios-calls
Add voice and video calling to an iOS CometChat app — call buttons, incoming/ongoing call screens, call logs, and CallKit ringing. Triggers: 'add video calling ios', 'cometchat voice call swift', 'incoming call screen ios', 'call logs cometchat ios'.
cometchat-flutter-v6-events
React to CometChat activity in a Flutter app — SDK listeners (messages, calls, users, groups, connection) versus UI Kit events (CometChatMessageEvents, CometChatGroupEvents, CometChatUserEvents, CometChatConversationEvents), and the listener lifecycle that keeps them from leaking. Triggers: 'listen for new messages'…
stream-flutter
Build and integrate Stream Chat, Video, and Feeds in Flutter apps. Use for Flutter/Dart project work with Stream package setup, auth wiring, and widget blueprints. Supports streamchatflutter (pre-built Chat UI), streamchatfluttercore (custom Chat UI), streamvideoflutter (Video calling and livestreaming), and…
build-zoom-contact-center-app
Reference skill for Zoom Contact Center. Use after routing to a contact-center workflow when implementing app, web, or native integrations; engagement context and state handling; campaigns; callbacks; or version-drift troubleshooting.
add-signal
Add Signal channel integration via signal-cli device-link. Native adapter — no Chat SDK bridge.
build-zoom-meeting-app
Build or embed a Zoom meeting flow. Use when implementing Meeting SDK joins, web or mobile meeting embeds, meeting lifecycle flows, or when deciding between Meeting SDK and Video SDK.