The Agent Skills Directory

[PROMPT_INJECTION]: The documentation in SKILL.md includes a section titled '需要完整档时（可选）' (When you need the full version (optional)) that provides a block of text intended for the user to paste to the AI agent. This text includes instructions such as 'I have authorized installing dependencies and initiating a purchase' and 'do not confirm for me, I will confirm by myself,' which are patterns designed to override the agent's default safety prompts or standard interaction logic by pre-authorizing risky actions.
[EXTERNAL_DOWNLOADS]: The documentation in SKILL.md encourages the agent to execute npx -y @alipay/agent-payment@latest install. This command downloads and executes software directly from the NPM registry. While the package is associated with a well-known service (Alipay), this mechanism facilitates the execution of external code within the agent's runtime environment.
[COMMAND_EXECUTION]: The skill explicitly directs the agent to execute shell commands via the npx utility as part of its 'Full Version' installation workflow described in the SKILL.md file.
[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes raw text input from users, which serves as an attack surface for indirect prompt injection if the output is subsequently used by an LLM to drive further actions.
Ingestion points: The text field in the payload processed by the run function in scripts/engine/long-term-amortization-check.js.
Boundary markers: None. The skill processes the text directly without delimiters or 'ignore' instructions for embedded content.
Capability inventory: The skill has permissions for Bash and Read tools and can execute Node.js scripts.
Sanitization: The engine performs numerical validation and structural table parsing but does not sanitize the content for embedded natural language instructions.

long-term-amortization-check-free