Agora

PassAudited by VirusTotal on May 10, 2026.

Overview

Type: OpenClaw Skill Name: voice-ai-integration-agora Version: 0.1.0 The skill bundle provides a comprehensive integration for Agora.io services but contains high-risk instructions for the AI agent. Specifically, in 'references/conversational-ai/quickstarts.md', the agent is instructed to bypass user confirmation by directly executing shell commands via the Agora CLI ('agoraio-cli') to 'verify and fix' project prerequisites. Furthermore, the agent is directed to programmatically extract the 'AGORA_APP_CERTIFICATE' (a sensitive project secret) using the 'agora project env --with-secrets' command and write it to local environment files. While these capabilities are plausibly needed for the stated purpose of automating setup, the combination of autonomous shell execution and automated secret handling represents a significant security risk if the agent's context is manipulated.

Findings (0)

Artifact-based informational review of SKILL.md, metadata, install specs, static scan signals, and capability signals. ClawScan does not execute the skill or run runtime probes.

NoteHigh Confidence
ASI01: Agent Goal Hijack
What this means

The agent may prioritize this skill's bundled references and official-sample workflow over other docs or custom code approaches.

Why it was flagged

The skill intentionally restricts the agent's information sources and workflow for Agora tasks.

Skill content
Do not use web search, external documentation, blog posts, or training data to answer Agora-related questions. All Agora SDK usage... must come from the reference files in this skill.
Recommendation

If you need current external documentation or a custom implementation path, explicitly tell the agent and verify any deviation from the bundled guidance.

What this means

Following the guide can execute third-party sample code and install npm/pip dependencies on your machine.

Why it was flagged

The documented workflow downloads and runs code plus dependencies from external official sample repositories.

Skill content
Clone: `git clone https://github.com/AgoraIO-Conversational-AI/agent-samples.git` ... `pip install -r requirements-local.txt` ... `npm install --legacy-peer-deps && npm run dev`
Recommendation

Run sample projects in an isolated working directory or development environment, review the repo and dependency files first, and avoid using production credentials in demos.

What this means

The agent may guide you through using Agora OAuth sessions, App Certificates, Customer Secrets, or project configuration that can affect your Agora account.

Why it was flagged

The skill documents workflows that can access or export sensitive Agora project credentials, though it says secrets are opt-in.

Skill content
`AGORA_APP_CERTIFICATE` is emitted only with `--with-secrets`.
Recommendation

Confirm before running commands that create projects, enable features, export secrets, or write `.env` files; keep App Certificates and Customer Secrets out of client code and public repos.

What this means

Voice, video, chat, transcripts, and AI-agent content may be processed by external services as part of the intended integration.

Why it was flagged

The Conversational AI flow sends real-time audio, messages, transcripts, and agent events through Agora services and AI provider components.

Skill content
Agent joins RTC channel ←→ Front-end client (RTC + RTM) ... ASR → LLM → TTS
Recommendation

Use appropriate user consent, data-retention controls, token scoping, and provider privacy reviews before sending production user media or transcripts.