The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Fetches documentation indices and page content from official domains doc.shengwang.cn and doc-mcp.shengwang.cn via curl in helper scripts.
[EXTERNAL_DOWNLOADS]: Clones official toolsets and sample applications from github.com/AgoraIO and gitee.com/agoraio-community repositories using git clone.
[COMMAND_EXECUTION]: Executes local bash scripts (fetch-docs.sh and fetch-doc-content.sh) to perform network-based documentation retrieval and file system updates.
[INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it retrieves and processes content from external documentation URIs.
Ingestion points: Documentation content fetched from the remote URI docs://default/... via doc-mcp.shengwang.cn.
Boundary markers: Implements a "locked quickstart state machine" that restricts file access and uses structured prompt templates for user interaction to maintain context integrity.
Capability inventory: Execution of local scripts, network downloads via curl, repository cloning via git, and multi-platform code generation.
Sanitization: Content fetched from the documentation server is ingested into the prompt context without explicit filtering or sanitization of the remote text.
[CREDENTIALS_SAFE]: Strongly enforces safe credential handling by instructing users to use environment variables (e.g., SHENGWANG_APP_ID, SHENGWANG_CUSTOMER_SECRET) and providing placeholder examples, while explicitly warning against hardcoding or committing secrets to version control.

voice-ai-integration