langchain-ai/deepagents · error · MissingCredentialsError

No credentials found for provider '{provider}'. Please set t

Error message

No credentials found for provider '{provider}'. Please set the {display_env} environment variable.

What it means

For non-Codex providers, deepagents-code raises MissingCredentialsError when no API key is found for the provider — neither the provider's known environment variable nor any stored credential. The message names the exact env var to set (resolved via get_credential_env_var), or a placeholder `<provider> API key` when no canonical env var is registered. The error carries `provider` and `env_var` so callers can offer targeted recovery.

Source

Thrown at libs/code/deepagents_code/config.py:5801

    if provider and provider not in IMPLICIT_AUTH_PROVIDERS:
        cred_status = has_provider_credentials(provider)
        if cred_status is False:
            from deepagents_code.model_config import MissingCredentialsError

            if provider == CODEX_PROVIDER:
                # No env var to set; point the user at `/auth` instead.
                msg = (
                    "Not signed in to ChatGPT. Run `/auth` and select "
                    "openai_codex to sign in with your ChatGPT account."
                )
                raise MissingCredentialsError(msg, provider=provider, env_var=None)
            env_var = get_credential_env_var(provider)
            display_env = env_var or f"<{provider} API key>"
            msg = (
                f"No credentials found for provider '{provider}'. "
                f"Please set the {display_env} environment variable."
            )
            raise MissingCredentialsError(msg, provider=provider, env_var=env_var)

    # Provider-specific kwargs (with per-model overrides)
    kwargs = _get_provider_kwargs(provider, model_name=model_name)

    # Compose under existing kwargs: profile < config.toml < --model-params
    # (applied below). The app's OpenRouter profile is stacked on top of the
    # built-in SDK profile so its `pre_init` (version check) and factory
    # (app attribution) compose into a single `apply_provider_profile` call.
    if provider:
        from deepagents.profiles.provider import apply_provider_profile

        if provider == "openrouter":
            _ensure_cli_openrouter_profile_registered()

        spec = f"{provider}:{model_name}" if model_name else provider
        try:
            kwargs = apply_provider_profile(spec, kwargs)
        except ModelConfigError:

View on GitHub (pinned to a1af029e6e)

Solutions

Export the env var named in the error message (e.g. `export ANTHROPIC_API_KEY=sk-...`) before launching dcode.
Run `/auth` inside dcode to store the key persistently instead of relying on the environment.
If the provider is wrong, correct the model spec / config.toml model entry to a provider you have credentials for.
For custom OpenAI-compatible endpoints, ensure the canonical env var for that provider is set (see PROVIDER_API_KEY_ENV).

Example fix

// before
$ dcode -m anthropic:claude-sonnet-4
// MissingCredentialsError: No credentials found for provider 'anthropic'...
// after
$ export ANTHROPIC_API_KEY=sk-ant-...
$ dcode -m anthropic:claude-sonnet-4

Defensive patterns

Strategy: validation

Validate before calling

import os
from deepagents_code.config import get_credential_env_var

def provider_credentials_present(provider: str) -> bool:
    env_var = get_credential_env_var(provider)
    return bool(env_var and os.environ.get(env_var))

if not provider_credentials_present("anthropic"):
    raise SystemExit("Set ANTHROPIC_API_KEY (or run /auth) before starting.")

Try / catch

try:
    model = build_model(provider, model_name)
except MissingCredentialsError as exc:
    if exc.env_var:
        print(f"Export {exc.env_var} or run /auth to store the key.")
    raise SystemExit(1)

Prevention

Export provider API keys in your shell profile or .env so every dcode session inherits them.
Prefer `/auth` for persistent credential storage that survives env changes.
In CI, verify required secret env vars exist before the job builds a model.
Double-check the provider part of model specs in config.toml matches a provider you configured.

When it happens

Trigger: Creating a chat model for a provider (e.g. `anthropic`, `openai`, `openrouter`) whose credential env var (e.g. ANTHROPIC_API_KEY) is unset in the process environment and no `/auth`-stored credential exists. Raised from the credential check in config.py when building the model.

Common situations: Running dcode in a fresh shell/container where API keys live in an rc-file that was not sourced; CI jobs missing secret env vars; renaming a model spec to a provider you never configured; a typo'd provider key in config.toml pointing at an unconfigured provider.

Understand the failure class

Background: "environment variable is not set" and "Missing keys in environment" errors: what missing required env var messages mean and how to fix them — this error's family across 28 libraries.

Related errors

AI-assisted analysis of langchain-ai/deepagents@a1af029e6e (2026-08-29). Data as JSON: /api/errors/1097ce6d4a268f2d. Report an issue: GitHub.