The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests data from local project files, including READMEs, CHANGELOGs, and git logs, to infer project status and minimize user questioning. This provides a potential surface for indirect prompt injection if those project files contain malicious instructions.
Ingestion points: The skill reads from developer-platform-context.md, git history, README files, and specification directories.
Boundary markers: While the skill suggests organizing the project context file into sections, it does not implement strict delimiters or specific instructions for the agent to ignore hidden directives within the analyzed project data.
Capability inventory: The skill performs routing and text generation, acting as the entry point for a collection that includes tools for CLI management and software publishing.
Sanitization: No explicit sanitization or validation logic is defined for the content extracted from the local environment.
[COMMAND_EXECUTION]: The skill utilizes local commands such as git log and file inventory processes to gather technical context about the repository and the project's development pace.
[EXTERNAL_DOWNLOADS]: The skill's routing logic and documentation reference various trusted package registries (such as npm, PyPI, and crates.io) and well-known technology platforms (including AWS, GitHub, and Slack) as part of its platform strategy guidance.

developer-platform-kickoff