LangSmith
Give the Sazabi agent the LangSmith CLI (langsmith) in its sandbox to inspect LLM traces, datasets, experiments, and evaluations.
About
This CLI connection makes the LangSmith CLI (langsmith) available in the Sazabi agent's sandbox. The agent can use it to inspect LLM traces, datasets, experiments, and evaluations. The tool is baked into the sandbox image; connecting it authenticates the already-installed CLI.
Prerequisites
- A Sazabi project.
- A LangSmith API key (
LANGSMITH_API_KEY), available from your LangSmith account settings. - Optionally, your LangSmith workspace ID (
LANGSMITH_WORKSPACE_ID) if you need to scope operations to a specific workspace.
Set up in the dashboard
Configure the sandbox in the dashboard under Settings > Sandbox CLIs.
Find LangSmith in the catalog
In Settings > Sandbox CLIs, find LangSmith under Browse sandbox CLIs and choose to connect it.
Provide credentials
Enter the required value: LANGSMITH_API_KEY — your LangSmith API key. Optionally, provide LANGSMITH_WORKSPACE_ID to scope operations to a specific workspace. Save to store them securely; Sazabi injects them into the sandbox when the agent runs.
Set up with the CLI
You can also configure the LangSmith CLI connection with the Sazabi CLI (installed and authenticated — see CLI reference).
sazabi sandbox-clis set --type langsmith --env LANGSMITH_API_KEY=<your-langsmith-key>To validate the credentials without saving, run the same command with test in place of set (the test command also requires the --env flags).
Verify
Ask the agent, in a thread, to run the langsmith CLI (a read-only or status command) and confirm it returns real output. If the command is not found, the CLI connection may not be enabled; if it runs unauthenticated or cannot connect, recheck the credentials.
Troubleshooting
Command not found — Confirm the LangSmith CLI connection is enabled for the project under Settings > Sandbox CLIs.
Authentication or connection fails — Recheck the credential/connection values under Settings > Sandbox CLIs.