AI API receipt · sgl-project/sglang

dots and Muse Glimmer detectors rewrite string-typed argument values (null becomes None, JSON-looking strings coerced)

sgl-project/sglang#43017 · upstream bug reproduced by running the issue's own reproduction

✓ Badgr rejects this response

A string parameter whose text was the word null arrived as JSON null (and the invalid boolean "yes" became false), as in the issue's detector-level reproduction.

Outcome contract verdict: tool call arguments do not match the tool schema. The attempt is recorded as failed (invalid_outcome) instead of a verified success, so it is not billed as a good answer and can fall back to another route.

What the upstream returned (replayed)

The output the bug produces, taken from the issue’s reproduction and fed to Badgr’s validators. Not a live provider call.

tool_calls: [{"type": "function", "function": {"name": "search", "arguments": "{\"query\": null, \"flag\": false}"}}]
tools: ["search"]
tool_choice: "required"

Badgr receipt

Written by Badgr’s receipt ledger and read back from GET /v1/receipts/replay-sglang-43017-string-args-rewritten.

{
  "request_id": "replay-sglang-43017-string-args-rewritten",
  "status": "failed",
  "endpoint": "/v1/chat/completions",
  "stream": false,
  "failure_stage": "outcome_contract",
  "error_code": "invalid_outcome",
  "error_message": "Response did not satisfy the outcome contract: tool call arguments do not match the tool schema",
  "retryable": true,
  "provider_mode": "replay",
  "routing_decision": "replayed_upstream_output",
  "created_at": "1791679879.072351"
}

Recorded 2026-10-11 in Badgr’s local development environment. Replay cases and verdicts are pinned by backend/test_ai_api_replay_cases.py.

← All AI API receipts