The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill fetches SKILL.md instruction files and JSON manifests from the author's GitHub repository (https://raw.githubusercontent.com/albertmartorell1975/android-ai-workflow-foundation/main/). These downloaded files are used to install additional agent capabilities into the project's .agents/skills/ directory.
[COMMAND_EXECUTION]: The skill directs the agent to execute shell commands, specifically git init and branch management operations, to establish a version control baseline for the new project.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-provided strings during the 'Stack Diagnosis' phase and interpolates them into governance templates for the project.
Ingestion points: User input for Project Name, Architecture, Dependency Injection tool, and Persistence library in SKILL.md Phase 3.
Boundary markers: The skill uses simple [PLACEHOLDER] markers in its templates but lacks robust delimiters or instructions to prevent users from injecting malicious instructions instead of data.
Capability inventory: The skill can perform file system writes, execute Git commands, and perform network requests to GitHub.
Sanitization: There is no evidence of input validation or escaping before user input is written into the project's governance files.

workflow-initializer