tirth8205/code-review-graph · critical · RuntimeError

Voyage API returned malformed indices (got {indices}, expect

Error message

Voyage API returned malformed indices (got {indices}, expected permutation of 0..{len(texts) - 1}) — refusing to misalign vectors.

What it means

The Voyage response items all have integer 'index' fields, but they are not a permutation of 0..len(texts)-1 (duplicates or out-of-range values). The provider sorts vectors by index before returning; a bad permutation would silently misalign vectors with texts, so it aborts.

Source

Thrown at code_review_graph/embeddings.py:756

                if "error" in response:
                    err = response["error"]
                    msg = err.get("message", "unknown") if isinstance(err, dict) else str(err)
                    raise RuntimeError(f"Voyage API error: {msg}")

                data = response.get("data", [])
                if not data:
                    raise RuntimeError("Voyage API returned empty data")

                any_has_index = any("index" in item for item in data)
                all_int_index = all(
                    isinstance(item.get("index"), int) for item in data
                )
                if all_int_index:
                    expected = set(range(len(texts)))
                    indices = [int(item["index"]) for item in data]
                    if len(set(indices)) != len(indices) or set(indices) != expected:
                        raise RuntimeError(
                            "Voyage API returned malformed indices "
                            f"(got {indices}, expected permutation of "
                            f"0..{len(texts) - 1}) — refusing to misalign vectors."
                        )
                    data = sorted(data, key=lambda item: int(item["index"]))
                elif not any_has_index:
                    if len(data) != len(texts):
                        raise RuntimeError(
                            f"Voyage API returned {len(data)} embeddings for "
                            f"{len(texts)} inputs with no index field — "
                            "refusing to misalign vectors."
                        )
                else:
                    raise RuntimeError(
                        "Voyage API returned mixed indexed/unindexed data — "
                        "refusing to misalign vectors."
                    )

View on GitHub (pinned to b58668751a)

Solutions

Retry the embed call — usually transient corruption
If using a proxy at CRG_VOYAGE_BASE_URL, bypass it and call Voyage directly
Reduce batch size (fewer texts per call) to avoid truncation
Capture the failing request payload and report the mismatch

Defensive patterns

Strategy: retry

Validate before calling

null

Type guard

def is_index_misalignment(exc: RuntimeError) -> bool:
    return "refusing to misalign vectors" in str(exc)

Try / catch

try:
    vecs = provider.embed(texts)
except RuntimeError as e:
    if "refusing to misalign vectors" in str(e):
        vecs = provider.embed(texts)  # retry once
    else:
        raise

Prevention

Avoid proxies that rewrite Voyage response bodies
Split large batches so responses stay small
Log and alert on any 'refusing to misalign' error — never ignore it

When it happens

Trigger: Calling embed() with multiple texts where the API returns duplicate/missing/garbage index values, e.g. truncated batch responses or a non-compliant OpenAI-compatible proxy in front of Voyage.

Common situations: Using an OpenAI-compatible gateway that mangles the index field, partial responses after network interruptions, or API behavior changes after batching changes.

Understand the failure class

Parsing and encoding errors: unexpected token, malformed input — why parsers reject input and how to find the real culprit.

Related errors

AI-assisted analysis of tirth8205/code-review-graph@b58668751a (2026-08-28). Data as JSON: /api/errors/c8764d290c686b93. Report an issue: GitHub.