Spandrift Analysis: head_trace.json
Spans: 15
Cost: $0.0152
Wall-clock: 235ms
Compute: 778ms
Per-Agent Rollup
| Agent | Cost | Wall-clock | Compute | LLM calls | Tokens (in/out) |
| FactChecker | $0.002075 | 86ms | 160ms | 1 | 350/120 |
| Orchestrator | $0.000000 | 235ms | 235ms | 0 | 0/0 |
| ResearchAgent | $0.006800 | 150ms | 256ms | 2 | 1.0k/420 |
| WriterAgent | $0.006325 | 74ms | 126ms | 1 | 850/420 |
Execution Waterfall & Concurrency
| Task / Span | Timeline | Duration | TTFT | Cost |
chat gpt-4o | | 31ms | 21ms | $0.002075 |
web_search | | 11ms | — | — |
chat gpt-4o | | 42ms | 21ms | $0.003400 |
web_search | | 10ms | — | — |
web_search | | 11ms | — | — |
ResearchAgent | | 75ms | — | — |
web_search | | 11ms | — | — |
web_search | | 11ms | — | — |
FactChecker | | 86ms | — | — |
chat gpt-4o | | 42ms | 21ms | $0.003400 |
web_search | | 11ms | — | — |
ResearchAgent | | 75ms | — | — |
chat gpt-4o | | 52ms | 31ms | $0.006325 |
WriterAgent | | 74ms | — | — |
Orchestrator | | 235ms | — | — |
⚠ Duplicate Calls
FactChecker called 4× with identical input (hash: 43635ec) wasted: $0.000000
ResearchAgent called 2× with identical input (hash: 377f6f9) wasted: $0.003400
ResearchAgent called 2× with identical input (hash: bdf540a) wasted: $0.000000
⚠ Retry/Loop Storms
chat gpt-4o: 4 calls with identical input (chain: bdcf538c → 002ec832 → 7c745c0f → 5e0ba9c7)
web_search: 6 calls with identical input (chain: c850c0c5 → 5f81509b → c3321927 → 6859e408)