The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill integrates WeChat messaging, creating an attack surface where untrusted data from external WeChat users is ingested into the agent's context.
Ingestion points: WeChat messages are received by the wechat-ilink-client adapter and passed to the agent (documented in SKILL.md).
Boundary markers: Absent. The instructions do not specify the use of delimiters or system prompts to ignore instructions within user messages.
Capability inventory: The agent can process these messages and potentially execute associated tools or logic defined in the agent's group configuration.
Sanitization: Absent. No specific sanitization or filtering of inbound message content is described in the skill configuration.
[COMMAND_EXECUTION]: The helper script scripts/wire-dm.ts executes local CLI commands using spawnSync to manage project configurations.
The script programmatically calls the project's own CLI tool (src/cli/client.ts) via pnpm exec tsx to list groups, wirings, and create new wiring entries.
[EXTERNAL_DOWNLOADS]: The skill installs an external dependency required for its primary functionality.
The wechat-ilink-client@0.1.0 package is installed from the public NPM registry to facilitate the connection to Tencent's iLink API.

add-wechat