Build with Mubit

Retrieve data

Use recall for answer-oriented retrieval and getContext for model-ready context assembly, with raw query only for explicit control.

Mubit exposes two default read patterns:

client.memory.recall() when you want Mubit to answer a question from memory
client.memory.context() (or run.context() inside a run) when you want a context block to send to your own model

Use raw control.query through client.raw.invoke when you need explicit control over low-level request fields or want to debug wire-level behavior.

Retrieval decision model

Need	Recommended method
Answer a question from stored memory	`client.memory.recall()`
Build a prompt-ready context block	`client.memory.context()` / `run.context()`
Inspect memory quality and gaps	`client.memory.diagnose()` and `client.memory.health()`
Explicit raw control over query payload	`client.raw.invoke("control.query", ...)`
Compatibility-only direct lanes	`client.raw.invoke("core.<op>", ...)` routes when explicitly enabled

Helper-first retrieval examples

retrieve_helper.py

def retrieve_helper(client, run_id: str, question: str):
    answer = client.memory.recall(
        question,
        entry_types=["fact", "lesson", "rule"],
        run_id=run_id,
    )
    context = client.memory.context(
        question,
        lane="legacy",
        mode="summary",
        max_token_budget=400,
        run_id=run_id,
    )
    return answer, context

lane="legacy" selects the /v2/control/context assembler with its mode values (full, summary, sections). Without it, client.memory.context(task, budget=...) returns the run-scoped ContextBlock the wrappers inject; see SDK helpers.

What the helper path buys you

client.memory.recall() uses the current control-plane query contract without you assembling the raw payload manually.
client.memory.context() gives you sectioned, cache-friendly context assembly with optional budget control.
client.memory.diagnose() and client.memory.health() help you understand why retrieval is weak before you start changing prompts blindly.

ℹ️Note

The recall() / query response includes a citations array — 0-based indices into evidence marking which items grounded final_answer (empty when the answer abstains or cites no specific evidence). Use them to render source attributions or audit answer grounding.

Failure modes and troubleshooting

Symptom	Root cause	Fix
Answer ignores an important lesson	Entry types or context budget too broad	Restrict `kinds` and use `client.memory.context()` budgeting
Retrieval feels noisy	You are mixing compatibility-only direct lanes with routed paths	Standardize one default read path per workflow
You need wire-level parity checks	Helper abstracts too much	Drop to `client.raw.invoke("control.query", ...)` for that case

Next steps

Inspect the helper and raw surface split at SDK methods.
Review the full public control surface at Control HTTP reference.