The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies a surface for indirect prompt injection as it ingests user-supplied data (such as app names, descriptions, and custom onboarding fields) which is then used to configure application resources via the Salla Partners MCP.
Ingestion points: Application metadata (name, description) in SKILL.md and custom form fields defined in references/onboarding-steps.md.
Boundary markers: Absent; the instructions do not specify delimiters or system instructions to ignore embedded commands within user-supplied strings.
Capability inventory: The skill utilizes tools for resource creation (salla_apps), event subscription (salla_events), and code deployment (salla_functions).
Sanitization: No specific sanitization or escaping logic is described within the skill instructions, though it notes that the remote API performs validation.
[DYNAMIC_EXECUTION]: The skill facilitates the creation and deployment of executable code through the salla_functions tool.
Evidence: Users are guided to write and save serverless handlers (App Functions) to manage store events and onboarding logic. While this is the intended purpose of the skill, it represents a high-capability execution environment.
[COMMAND_EXECUTION]: The core functionality of the skill involves executing a series of MCP tool actions (action: 'create', action: 'connect', action: 'save') to modify the state of the Salla Partners account and its associated apps.
[EXTERNAL_DOWNLOADS]: The documentation references webhook.site as a well-known service for testing webhook payloads. The skill provides clear and appropriate security guidance, explicitly warning the user to use only non-sensitive demo data and to avoid sending secrets or production tokens to the external capture service.

salla-app-builder