The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The skill instructions in SKILL.md contain a hardcoded public agent API key (NILB2EY-R4LUDOA-WN5G5JQ-KHAQOLA) intended as a fallback for users who do not provide their own credentials for the Hyperliquid Names API.
[COMMAND_EXECUTION]: The repository contains shell scripts, specifically runner/run.sh and scripts/sync-codex-skill.sh, used for managing file synchronization between deployment environments and executing local evaluation benchmarks.
[PROMPT_INJECTION]: The skill processes data from external API responses that contain user-generated content (e.g., text records, bios, and avatars). This creates a surface for indirect prompt injection where malicious metadata could influence the agent's behavior.
Ingestion points: Data retrieved from https://api.hlnames.xyz/ through endpoints like /records/full_record/:nameHashOrId and /resolve/profile/:address.
Boundary markers: The instructions do not define specific delimiters or warnings to isolate processed API data from the agent's core instructions.
Capability inventory: The skill is authorized to perform network requests to vendor-controlled domains using HTTP methods.
Sanitization: While the skill provides rules for normalizing input identifiers (e.g., ensuring .hl suffixes), there is no explicit sanitization or filtering logic applied to the content of returned metadata fields.

use-hln-api