Shared env sets
A shared env set is a reusable collection of environment variables you can attach to
multiple agents. Instead of pasting the same OPENAI_API_KEY or SLACK_BOT_TOKEN into
every agent, you define it once in a shared env set and reference it from each agent that
needs it.
Why use shared env sets
- Avoid duplication — define a secret once, reuse it across agents.
- Rotate in one place — update the value in the shared env set and every attached agent picks it up.
- Keep secrets consistent — no risk of one agent using an outdated key.
Creating a shared env set
- Open the Shared env sets page in Hermeum.
- Select New shared env set.
- Give it a name (for example,
shared-openai-key) and an optional description. - Add the environment variables you want to share. Each variable has a name and a value — mark credentials as sensitive so they're stored securely.
Attaching to agents
When creating or editing an agent, reference a shared env set from the agent's env section. The variables in the set are merged with the agent's own env vars at runtime.
Example
env:
- name: LOG_LEVEL
value: debug
sharedEnvSets:
- envset-01H8X9F7...
In this example, LOG_LEVEL is specific to the agent, while everything in the
referenced shared env set (for instance, OPENAI_API_KEY) is pulled in from the set.
Next steps
- Creating an agent — the full agent model, including env.
- Messaging platforms — platforms that commonly need shared credentials like bot tokens.