The Agent Skills Directory

[COMMAND_EXECUTION]: The ToolBridge implementation described in SKILL.md explicitly maps a bash interface to the system's exec command, allowing for arbitrary shell execution.
Evidence: bash: async (cmd) => exec(cmd) in the ToolBridge class definition within SKILL.md.
[PROMPT_INJECTION]: The skill architecture creates a significant surface for indirect prompt injection by combining data ingestion tools with high-privilege execution tools.
Ingestion points: The toolRegistry includes webfetch and websearch for retrieving external data.
Boundary markers: No explicit delimiters or instructions to ignore embedded commands are present in the provided summarization or tool-call logic.
Capability inventory: The skill facilitates bash (shell access), write (file creation), and edit (file modification).
Sanitization: There is no evidence of sanitization or validation of data retrieved from external URLs before it is processed by the agent or passed to other tools.

token-optimizer