The Agent Skills Directory

[PROMPT_INJECTION]: The skill processes untrusted data from git logs and internal documents that could contain malicious instructions designed to influence the agent's output.
Ingestion points: Reads git history (git log) and project files in production/releases/, production/sprints/, and design/balance/.
Boundary markers: No explicit delimiters are specified to separate ingested content from the agent's processing instructions.
Capability inventory: The skill possesses Write access to the filesystem and Bash access for command execution.
Sanitization: The instructions guide the agent to remove technical jargon and internal IDs but do not include specific sanitization logic for prompt-based attacks.
[COMMAND_EXECUTION]: The skill uses the Bash tool to execute git log to retrieve version history, which is a legitimate use case for its primary function.
[SAFE]: No network exfiltration or use of hardcoded credentials was identified in the skill.

patch-notes