The Agent Skills Directory

[COMMAND_EXECUTION]: The skill utilizes a bash helper script scripts/crawlora.sh to execute curl commands for API interaction.
The script enforces a fixed base URL (https://api.crawlora.net/api/v1), preventing the agent from being redirected to malicious endpoints.
It implements a strict whitelist for API paths using regular expressions and string matching to ensure only documented anime/manga research endpoints are accessed.
The script securely handles the CRAWLORA_API_KEY by writing it to a temporary, restricted-permission curl configuration file (chmod 600) rather than passing it as a command-line argument, protecting it from exposure in process lists.
User-supplied query parameters are safely handled using curl's --data-urlencode functionality.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from an external API, which is an inherent surface for indirect prompt injection.
Ingestion points: External data enters the agent context via JSON responses from api.crawlora.net inside the scripts/crawlora.sh script.
Boundary markers: No explicit instructions are provided to the model to ignore instructions embedded within the fetched data.
Capability inventory: The skill possesses network capabilities via curl and file-read capabilities via standard shell operations in its helper script.
Sanitization: The script sanitizes outgoing query parameters via URL-encoding. The documentation recommends the use of jq for handling the resulting JSON data, which promotes structured data processing over raw text interpretation.

anime-manga-research