The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted content from the local repository (e.g., GLOSSARY.md, docs/agents/domain.md, CONTEXT.md) to use as context for generating tests and code.
Ingestion points: SKILL.md explicitly instructs the agent to read various documentation and configuration files from the project root and docs/ folder.
Boundary markers: No explicit boundary markers or "ignore embedded instructions" warnings are provided for the ingested file content.
Capability inventory: The skill allows the agent to modify code and tests, providing a significant capability tier.
Sanitization: No sanitization or validation of the ingested text is performed before it is used to influence the agent's behavior.
[SAFE]: No other malicious patterns, such as obfuscation, credential exfiltration, or unauthorized network access, were detected. The file access is restricted to documentation files relevant to the skill's primary TDD purpose.

tdd