The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses a helper script scripts/crawlora.sh to interface with the Crawlora API. This script implements several security best practices: it uses a whitelist to restrict execution to specific travel-related routes, validates the API key format to prevent shell injection, and uses a temporary configuration file to pass credentials to curl without exposing them in the process list. It also explicitly blocks the @ character in query parameters to prevent curl from accidentally reading and disclosing local files.
[EXTERNAL_DOWNLOADS]: The skill fetches data from api.crawlora.net. These network operations are strictly limited by the helper script to a set of predefined travel research endpoints (Airbnb, Agoda, Hotels.com, Trip.com, and Vrbo). The target domain is owned by the skill's author, crawlora-org.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process public listing data, including guest reviews from travel platforms. This external content is untrusted and could potentially contain adversarial instructions aimed at influencing the agent's logic.
Ingestion points: API responses from api.crawlora.net containing user-generated content such as host bios and property reviews (identified in SKILL.md and reference/endpoints.md).
Boundary markers: The skill instructions advise the agent to keep data fields like nightly rates and platform identity separate and to report the collection scope, which helps maintain data integrity, though explicit prompt delimiters for the final output are not defined.
Capability inventory: The skill possesses the capability to perform network requests to a specific travel API via the provided bash script. It does not have access to sensitive system files or broader network capabilities.
Sanitization: The helper script scripts/crawlora.sh performs robust sanitization of inputs (query parameters and request bodies) to prevent command injection at the shell level, but the skill does not include specific logic to sanitize the content of retrieved reviews for potential prompt injection before the agent processes them.

travel-accommodation-research