The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill explicitly instructs the agent to read and prioritize instructions from a project-local file: .ai-factory/skill-context/aif-roadmap/SKILL.md.
Ingestion points: The skill reads project-controlled files including .ai-factory/config.yaml, .ai-factory/DESCRIPTION.md, and specifically .ai-factory/skill-context/aif-roadmap/SKILL.md to resolve logic and rules.
Boundary markers: None. The instructions explicitly state that project-specific rules should "win" and "override" the built-in instructions: "When a skill-context rule conflicts with a general rule written in this SKILL.md, the skill-context rule wins".
Capability inventory: The skill uses Write and Edit tools to modify files and Bash(git *) to interact with the repository's history.
Sanitization: The skill lacks sanitization or validation of the rules ingested from the codebase, treating them as authoritative overrides.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute git log commands to analyze project history. While the tool is restricted to the git binary, this constitutes automated subprocess execution based on logic that can be influenced by project metadata.

aif-roadmap