This repository currently contains automation configuration for the project, including Crowdin synchronization workflows.
Linear issue SAN-11 reported that /hermes-agent failed while loading the
hermes-agent skill because calls to the Hugging Face router returned HTTP
402 Payment Required for Qwen/Qwen3-Coder-Next.
The error means the Hugging Face account or organization used by the agent has exhausted its included Inference Providers credits. This is an environment or billing configuration issue, not a repository code failure.
To restore the Hermes-agent session:
- Add prepaid credits or enable the required paid billing plan for the Hugging Face account or organization used by the agent token.
- If the token should bill an organization, configure the agent runtime to send
the appropriate Hugging Face billing organization header, such as
X-HF-Bill-To, when supported by the provider path. - Alternatively, switch the agent runtime to another configured model provider with available quota.
- Re-run
/hermes-agentafter updating billing or provider configuration.
Keep agent provider credentials and billing headers in the runtime environment or secret store. Do not commit tokens, API keys, or billing secrets to this repository.