The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes the signet CLI tool for initial setup and uses the find command to scan specific home directory paths (such as ~/.openclaw and ~/.moltbot) to audit existing agent configurations for integration health.
[DATA_EXFILTRATION]: Conducts health and status checks by sending network requests via curl to the local Signet daemon at localhost:3850. These operations are limited to the local machine and do not target external domains.
[PROMPT_INJECTION]: The skill implements a functional surface for indirect prompt injection by allowing user-defined 'Hard Rules' to be written into the AGENTS.md configuration file. 1. Ingestion points: User-provided input for custom instructions and rules during Step 4 of the onboarding process. 2. Boundary markers: The template directly interpolates the {{custom_instructions}} variable without additional delimiters or instructions to ignore embedded commands. 3. Capability inventory: The skill has the ability to write to the filesystem, execute local CLI tools, and perform network requests to the local daemon. 4. Sanitization: No sanitization or validation is applied to user-provided strings before they are persisted to the configuration files.

onboarding