The Agent Skills Directory

[DATA_EXPOSURE]: The skill communicates with api.crawlora.net. This domain is the official infrastructure for the skill author, crawlora-org. The script properly handles the CRAWLORA_API_KEY by passing it to curl via a temporary configuration file with restricted permissions (chmod 600), preventing the secret from appearing in process monitoring tools like ps.
[COMMAND_EXECUTION]: The helper script scripts/crawlora.sh implements strict input validation. It enforces a whitelist of specific API endpoints, validates the HTTP method, and sanitizes the URL path to prevent path traversal or injection.
[DATA_EXPOSURE]: The script explicitly checks for and rejects the @ character in query parameters. This is a security best practice for curl-based wrappers to prevent an attacker from tricking the tool into reading and uploading local files via curl's file-reading shorthand.
[INDIRECT_PROMPT_INJECTION]: The skill processes data from the Crawlora API, which creates a theoretical attack surface for indirect prompt injection if the API were to return malicious instructions.
Ingestion points: Results returned by scripts/crawlora.sh from the Crawlora API.
Boundary markers: Absent.
Capability inventory: Network access via curl to api.crawlora.net for APK teardown analysis.
Sanitization: The helper script performs rigorous validation on the API path and query arguments to prevent unauthorized tool use.

apk-teardown-research