The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructions include a process for executing a curl command retrieved from the Testany API (via testany_log_sign or testany_get_dry_run_log) to fetch logs. While this involves dynamic command execution using eval, the skill mandates a strict verification process: checking that the URL uses HTTPS, belongs to the vendor's trusted domains (testany.io or testany.com.cn), and does not contain dangerous shell operators like pipes, redirects, or command substitutions.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze test logs, which are external data sources that could theoretically contain malicious instructions.
Ingestion points: Logs are fetched via curl from execution and dry-run endpoints.
Boundary markers: None explicitly defined for log content processing.
Capability inventory: The skill uses tools to query workspace status and execution details, but does not have arbitrary file-write or network-send capabilities beyond fetching logs from specified domains.
Sanitization: The skill implements domain and parameter validation for the retrieval command itself, though the content of the logs is analyzed by the LLM for diagnostic purposes.

testany-debug