BerriAI/litellm · error · AgentCardDiscoveryError

langgraph_platform discovery requires params.assistant_id

Error message

langgraph_platform discovery requires params.assistant_id

What it means

Validation when building A2A discovery paths for LangGraph Platform: params (or params.assistant_id) is missing, but LangGraph serves agent cards per assistant, so the assistant_id query parameter is mandatory.

Source

Thrown at litellm/proxy/a2a/discovery.py:69

def _normalize_base_url(base_url: str) -> str:
    return base_url.rstrip("/")


def _build_langgraph_platform_paths(
    params: dict[str, Any] | None,
) -> tuple[str, ...]:
    """Build the paths to try for LangGraph Platform discovery.

    LangGraph serves the card at ``/.well-known/agent-card.json`` with the
    ``assistant_id`` carried as a query parameter. We still try the other
    A2A path variants (with the same query string appended) so we degrade
    gracefully if a deployment uses an older spec name.
    """
    assistant_id: Final = (params or {}).get("assistant_id")
    if not assistant_id:
        raise AgentCardDiscoveryError("langgraph_platform discovery requires params.assistant_id")
    query: Final = urlencode({"assistant_id": str(assistant_id)})
    return tuple(f"{path}?{query}" for path in AGENT_CARD_WELL_KNOWN_PATHS)


def _paths_for_mode(mode: DiscoveryMode, params: dict[str, Any] | None) -> tuple[str, ...]:
    if mode == DiscoveryMode.WELL_KNOWN_FALLBACK:
        return AGENT_CARD_WELL_KNOWN_PATHS
    if mode == DiscoveryMode.LANGGRAPH_PLATFORM:
        return _build_langgraph_platform_paths(params)
    raise AgentCardDiscoveryError(f"unsupported discovery_mode: {mode}")


async def fetch_well_known_card(
    base_url: str,
    *,
    discovery_mode: DiscoveryMode = DiscoveryMode.WELL_KNOWN_FALLBACK,
    params: dict[str, Any] | None = None,
    timeout: float = DEFAULT_DISCOVERY_TIMEOUT_SECONDS,

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Include params.assistant_id in the discovery request.

Example fix

params={'assistant_id': 'asst_...'}

When it happens

Trigger: Thrown at litellm/proxy/a2a/discovery.py:69 when the library encounters an invalid state.

Common situations: See trigger scenarios.

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/5e688f9e3529323b. Report an issue: GitHub.