The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill identifies and processes project-level overrides from context files which are given priority over its internal instructions. This creates a vulnerability surface where malicious rules could be introduced via these files.
Ingestion points: The agent reads project context from .ai-factory/DESCRIPTION.md and rules from .ai-factory/skill-context/aif-architecture/SKILL.md (Step 0).
Boundary markers: There are no markers or instructions used to delimit untrusted data or prevent the agent from following malicious instructions found in those files.
Capability inventory: The skill possesses Write permissions to multiple files (ARCHITECTURE.md, DESCRIPTION.md, AGENTS.md) and uses Bash for directory creation.
Sanitization: The skill lacks mechanisms to sanitize or validate the content of the project files before they are incorporated into its decision-making logic.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute mkdir commands, enabling the creation of necessary directory structures for generated artifacts (Step 2).

aif-architecture