The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill provides workflows for handling sensitive cryptographic private keys. It supports passing these keys to the altllm CLI via environment variables (ALTLLM_WALLET_PRIVATE_KEY) or directly as command-line arguments using the --private-key flag. The documentation explicitly acknowledges that direct argument usage is unsafe and requires a bypass flag (--allow-unsafe-private-key-argv) because of the risk that keys could leak into shell history or system process listings.
[COMMAND_EXECUTION]: The skill is designed to execute shell commands using node dist/cli.js to create payment links and settle cryptocurrency transactions. These operations involve high-risk execution as they handle financial assets and are driven by parameters provided by the agent.
[PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface.
Ingestion points: The skill ingests untrusted data from the AltLLM Portal API, specifically via the GET /api/billing/payment-links endpoint and payment creation responses.
Boundary markers: No explicit delimiters or instructions to ignore instructions within the API data are provided to the agent.
Capability inventory: The skill utilizes shell command execution via node dist/cli.js across its primary workflows.
Sanitization: The instructions do not describe any sanitization or validation of external data (such as discount codes, metadata, or currency identifiers) before it is used to construct CLI commands.

altllm-portal-payments