WorkflowBack to dashboard

Flight Reservation Modification Flight Cancellation Refund Other Cabin Class Upgrade Downgrade New Flight Booking Booking Conflict Resolution Flight Delay Compensation

Metric scope

Workflow metrics are projected; evidence stays tied to analyzed threads

Projected workflow runs63 / moThis workflow represented 19 of 300 analyzed threads.

Analyzed workflow sample19 threadsFindings, recommendations, and evidence cards are still anchored to the normalized workflow sample.

Projection factor3.3xApplied to this workflow's spend, savings, runs, and token totals. Confidence: medium.

Source pool996 sessionsThe full source-pool population used by the dashboard projection.

derived-flight-delay-compensation · Based on 19 threads · medium confidence

Flight Delay Compensation

Derived primarily from user-authored prompts across a 300-thread slice. Full-slice prompt clustering ran on every thread, and Claude consolidated the major workflow types from cluster exemplars because the slice exceeds the non-sampling threshold.

Projected spend / mo$1.10sample $0.33

Projected savings / mo$0.27sample $0.08 · Could cut spend by ~25%

Projected runs / mo63sample 19

Projected total tokens284.8Kavg 4.5K per run

Projected input / output54.2K / 230.5K

This workflow

Token and spend trend

8 hour buckets

Dec 4 11AM

Dec 4 1PM

Dec 17 4PM

Dec 17 5PM

Dec 17 6PM

Dec 17 7PM

Dec 17 8PM

Dec 17 9PM

Input tokensOutput tokens

Model mix

Tokens and spend by model

4 models

Tokens by model

input + output

Kimi-K252.1K tokens

98 calls34.6% of total

Claude Opus 4.539.2K tokens

67 calls26.1% of total

gemini-3-pro-preview30.4K tokens

85 calls20.2% of total

GPT-5.228.6K tokens

89 calls19% of total

Spend by model

estimated cost

Claude Opus 4.5$0.77

68.7% of total

GPT-5.2$0.22

19.8% of total

Kimi-K2$0.09

8.5% of total

gemini-3-pro-preview$0.03

3% of total

Opportunities

2 opportunities for this workflow

$0.27 projected

Tool misuse

Failed benchmark outcomes are still paying the full workflow cost

The imported outcome labels show a high failure rate after the workflow has already spent tokens and tool calls, which points to missing early exits or weak preflight checks.

$0.13projected / month savedsample $0.04/mo

high riskmedium confidence

Recommended changes

Compare passing and failing traces for this workflow and add an early gate before the expensive tool loop starts.
Use the imported outcome label as an evaluation dimension so regressions are ranked by wasted spend, not just by raw failure count.

Evidence (1)

StepImported failing outcome

Imported benchmark outcome ended with failure

Tool misuse

Tool loops are dense enough to need batching or early stopping

tool dominates repeated tool activity, so the workflow is likely doing incremental calls where batching, caching, or tighter stop conditions would reduce churn.

$0.13projected / month savedsample $0.04/mo

medium riskmedium confidence

Recommended changes

Batch or cache repeated tool calls where the inputs overlap across adjacent steps.
Add a per-run tool budget and stop condition so failed runs do not keep exploring after the likely answer is already unreachable.

Prompt composition

Input token breakdown

16.3K tokens

user16.3K · 100%

Tool signals

How this workflow runs

Retries0

How often steps had to re-run.

Delegated subtasks0

Tasks handed off to sub-agents during the workflow.

Documents retrieved0

Total documents pulled in across all tool calls.

Median step latency0 ms

Typical time each step takes to finish.

Stage order

Typical workflow path

7 steps

RespondKimi-K2
Respond step in the workflow.
Latency unavailable313 tok avg
2
Loop×50
Loop: plan → tool — repeats 50 times.
Latency unavailable13.4K tok avg
1. 1
  PlanKimi-K2
  Plan the next steps in the workflow.
  Latency unavailable116 tok avg
2. 2
  Tooltool
  Tool step in the workflow.
  Latency unavailable152 tok avg
RespondKimi-K2
Respond step in the workflow.
Latency unavailable313 tok avg
4
Loop×15
Loop: plan → tool — repeats 15 times.
Latency unavailable4K tok avg
1. 1
  PlanKimi-K2
  Plan the next steps in the workflow.
  Latency unavailable116 tok avg
2. 2
  Tooltool
  Tool step in the workflow.
  Latency unavailable152 tok avg
RespondKimi-K2
Respond step in the workflow.
Latency unavailable313 tok avg
Tooltool
Tool step in the workflow.
Latency unavailable152 tok avg
Verify
Verify step in the workflow.
Latency unavailable119 tok avg

Threads

Pick a thread to see what happened

19 threads

Cost per run$0.05

Monthly runs1

Monthly cost$0.05

Operation path35 named tool/models

1
RespondRespond
claude-opus-4-5claude-opus-4-5
7 tok · $0.00
2
RespondRespond
claude-opus-4-5claude-opus-4-5
130 tok · $0.00
3
PlanPlan
claude-opus-4-5claude-opus-4-5
53 tok · $0.00
4
toolTool
Run tooltool
300 tok
5
PlanPlan
claude-opus-4-5claude-opus-4-5
108 tok · $0.00
6
toolTool
Run tooltool
202 tok
7
PlanPlan
claude-opus-4-5claude-opus-4-5
77 tok · $0.00
8
toolTool
Run tooltool
219 tok
9
PlanPlan
claude-opus-4-5claude-opus-4-5
180 tok · $0.00
10
toolTool
Run tooltool
14 tok
11
PlanPlan
claude-opus-4-5claude-opus-4-5
113 tok · $0.00
12
toolTool
Run tooltool
1.3K tok
13
PlanPlan
claude-opus-4-5claude-opus-4-5
102 tok · $0.00
14
toolTool
Run tooltool
29 tok
15
RespondRespond
claude-opus-4-5claude-opus-4-5
294 tok · $0.01
16
PlanPlan
claude-opus-4-5claude-opus-4-5
108 tok · $0.00
17
toolTool
Run tooltool
6 tok
18
RespondRespond
claude-opus-4-5claude-opus-4-5
216 tok · $0.00
19
PlanPlan
claude-opus-4-5claude-opus-4-5
171 tok · $0.00
20
toolTool
Run tooltool
11 tok
21
RespondRespond
claude-opus-4-5claude-opus-4-5
290 tok · $0.01
22
PlanPlan
claude-opus-4-5claude-opus-4-5
41 tok · $0.00
23
toolTool
Run tooltool
8 tok
24
PlanPlan
claude-opus-4-5claude-opus-4-5
38 tok · $0.00
25
toolTool
Run tooltool
5 tok
26
PlanPlan
claude-opus-4-5claude-opus-4-5
138 tok · $0.00
27
toolTool
Run tooltool
6 tok
28
PlanPlan
claude-opus-4-5claude-opus-4-5
56 tok · $0.00
29
toolTool
Run tooltool
56 tok
30
PlanPlan
claude-opus-4-5claude-opus-4-5
58 tok · $0.00
31
toolTool
Run tooltool
15 tok
32
PlanPlan
claude-opus-4-5claude-opus-4-5
138 tok · $0.00
33
toolTool
Run tooltool
41 tok
34
RespondRespond
claude-opus-4-5claude-opus-4-5
160 tok · $0.00
35
dataset evaluationTool
Run dataset_evaluationdataset_evaluation
100 tok
36
Imported benchmark outcomeVerify
Imported benchmark outcome
124 tok

The old plan/tool string was the normalized span order. Rows above use imported operation records; when a tool name is missing, the source only provided the normalized stage and operation label.

Snapshots

full_transcriptSnapshot 1 · imported

Hi! How can I help you today? Hi, I’m calling because I’m really frustrated about my last flight—it was delayed for hours and it messed up all my plans. Can you help me with this?…