The Agent Skills Directory

[COMMAND_EXECUTION]: The skill relies on the execution of multiple shell commands via the @insforge/cli tool to perform diagnostics and system inspections.
[CREDENTIALS_UNSAFE]: Explicitly provides instructions and commands to retrieve sensitive project secrets, including npx @insforge/cli secrets get <key> and npx @insforge/cli secrets list --all.
[DATA_EXFILTRATION]: Facilitates access to sensitive internal data including source code (npx @insforge/cli functions code <slug>), arbitrary database records through SQL queries (npx @insforge/cli db query "<sql>"), and comprehensive backend logs.
[EXTERNAL_DOWNLOADS]: Fetches and executes the @insforge/cli package from the NPM registry using npx. This package is associated with the skill's authoring organization.
[PROMPT_INJECTION]: Contains an indirect prompt injection surface through the ingestion of untrusted data:
Ingestion points: User-provided problem descriptions in the npx @insforge/cli diagnose --ai "<issue description>" command (SKILL.md), as well as system logs and source code retrieved during diagnostic steps.
Boundary markers: None identified to separate instructions from processed data.
Capability inventory: The skill has extensive capabilities including shell command execution, secret retrieval, and database access.
Sanitization: No sanitization or validation steps are described for the data processed by the agent.

insforge-debug