The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill retrieves and processes external data from the Appllama library, including app profiles and screen metadata. This data is used to inform the agent's design decisions and implementation steps, creating a potential vector for instructions embedded in external metadata to influence the agent.\n
Ingestion points: The tools search_apps, get_app, list_app_screens, search_screens, get_screen, list_my_boards, and get_board retrieve data from the mcp.appllama.io endpoint (as documented in SKILL.md).\n
Boundary markers: The instructions do not define specific delimiters or "ignore instructions" tags for data retrieved from the research tools.\n
Capability inventory: The skill records findings to the local file system in the research/ directory and guides the creation and simulation of application code as detailed in references/build-from-scratch.md.\n
Sanitization: No explicit sanitization or validation of the retrieved metadata is performed before it is processed by the agent.\n- [EXTERNAL_DOWNLOADS]: The skill configuration in agents/openai.yaml specifies a dependency on a Model Context Protocol (MCP) server at https://mcp.appllama.io/mcp. This connection is used to fetch design research data. The resource is owned by the authoring organization.

appllama-usage