The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses shell commands for directory management and profile resetting. These are scoped within the configuration:
Shell access is restricted to mkdir for creating directories and rm for deleting the specific profile file located at ~/.awase/profile.json.
[PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection due to its core functionality of analyzing session code to create exercises.
Ingestion points: The skill reads active session code (SKILL.md, Step 2) to identify 'Surface' and 'Underlying' concepts for training.
Boundary markers: No explicit boundary markers or delimiters are specified when the extracted concepts are used to generate exercise prompts.
Capability inventory: The skill has access to Read, Write, and restricted Bash tools to modify the file system and manage its profile.
Sanitization: The instructions include a robust security section requiring the agent to redact credentials, API keys, and tokens from both the exercise output and the stored profile before processing.

awase