abhigyanpatwari/GitNexus · error · SandboxError

transcript_artifacts exceeds the global evidence limit of {M

Error message

transcript_artifacts exceeds the global evidence limit of {MAX_TRANSCRIPT_ARTIFACTS}

What it means

Raised by `_preflight_transcript_artifacts` when the running total of transcript artifacts across ALL evidence rows exceeds `MAX_TRANSCRIPT_ARTIFACTS` (=24 = 12 rows x 2). This bounds the total evidence a proposer can read.

Source

Thrown at eval/workflow_bench/evolve.py:374


def _preflight_transcript_artifacts(evidence: list[dict[str, Any]]) -> list[list[Any]]:
    """Bound every transcript reference before any evidence file is read."""

    artifacts_by_row: list[list[Any]] = []
    seen_paths: set[str] = set()
    total = 0
    for artifacts_row in evidence:
        artifacts = artifacts_row.get("transcript_artifacts", [])
        if not isinstance(artifacts, list):
            raise SandboxError("transcript_artifacts must be a list")
        if len(artifacts) > MAX_TRANSCRIPT_ARTIFACTS_PER_ROW:
            raise SandboxError(
                f"transcript_artifacts exceeds the per-row session limit of {MAX_TRANSCRIPT_ARTIFACTS_PER_ROW}"
            )
        total += len(artifacts)
        if total > MAX_TRANSCRIPT_ARTIFACTS:
            raise SandboxError(f"transcript_artifacts exceeds the global evidence limit of {MAX_TRANSCRIPT_ARTIFACTS}")
        for artifact in artifacts:
            relative, _, _ = _transcript_artifact_metadata(artifact)
            normalized = _normalized_transcript_artifact_path(relative)
            if normalized in seen_paths:
                raise SandboxError(f"duplicate transcript artifact path: {normalized}")
            seen_paths.add(normalized)
        artifacts_by_row.append(artifacts)
    return artifacts_by_row


def _bound_transcript_artifact(root: Path, metadata: Any) -> str:
    relative, expected_digest, expected_size = _transcript_artifact_metadata(metadata)

    path = _results_artifact_path(root, relative, transcript=True)
    try:
        before = path.lstat()
    except OSError as exc:
        raise SandboxError(f"transcript artifact is unavailable: {path}: {exc}") from exc

View on GitHub (pinned to d540b00184)

Solutions

Let evidence flow only through select_evidence(), which caps rows at MAX_EVIDENCE_ROWS.
Reduce the per-row transcript count or drop the lowest-value rows so the total is <=24.
If you raised MAX_EVIDENCE_ROWS, raise MAX_TRANSCRIPT_ARTIFACTS to match (rows x per-row).

Defensive patterns

Strategy: validation

Validate before calling

total = sum(len(r.get('transcript_artifacts', [])) for r in evidence)
if total > 24:
    raise ValueError(f'transcript_artifacts total {total} exceeds global cap 24')

Prevention

Always pass evidence through select_evidence; do not concatenate extra rows by hand.
Keep MAX_TRANSCRIPT_ARTIFACTS == MAX_EVIDENCE_ROWS * MAX_TRANSCRIPT_ARTIFACTS_PER_ROW if you change either.

When it happens

Trigger: `proposer_evidence_entries` receives more than 24 transcript_artifacts total across the selected evidence rows. Since `select_evidence` already caps rows at MAX_EVIDENCE_ROWS (12), this only fires when rows are over-packed or the cap constants were lowered.

Common situations: Manually injecting extra rows beyond what select_evidence returns, or lowering MAX_EVIDENCE_ROWS without lowering the per-row cap proportionally so the product still fits.

Related errors

AI-assisted analysis of abhigyanpatwari/GitNexus@d540b00184 (2026-08-12). Data as JSON: /api/errors/2b0d6c193a5ea50b. Report an issue: GitHub.