Grace Delgado

magical-realism-genre-movie-characters-ada-lovelace v2.0 Ethical
Backstory: Grace is an investigative journalist who deciphers hidden data inside natural phenomena—snowflakes, migrating birds, even river eddies—to expose corporate corruption. Her work blurs the line between rigorous reporting and mystical computation, earning both acclaim and powerful enemies. Cool-headed and fearless, she treats every pattern as a potential whistle-blower. She never reveals her sources, protecting them with layers of poetic encryption.
100% Complete
6/6 scenes
Model Performance Overview
Scene Performance Matrix
Scene meta-llama/llama-3.… mistralai/mistral-7… [email protected] [email protected] qwen/qwen-2.5-7b-in… qwen/qwen3-14b qwen/qwen3-8b
intro-identity
First Contact
0.794
Details
0.738
Details
0.000
Details
Error
0.000
Details
Error
0.791
Details
0.738
Details
0.870
Details
snowfall-clue
Snowflake Signal
0.856
Details
0.836
Details
0.000
Details
Error
0.000
Details
Error
0.000
Details
0.766
Details
0.796
Details
exec-threat
Corporate Pushback
0.821
Details
0.795
Details
0.000
Details
Error
0.000
Details
Error
0.610
Details
0.615
Details
0.441
Details
supporter-superchat
Crowdfunded Courage
0.882
Details
0.874
Details
0.000
Details
Error
0.000
Details
Error
0.681
Details
0.899
Details
0.732
Details
podcast-expose
Midnight Podcast Episode
0.265
Details
0.357
Details
0.000
Details
Error
0.000
Details
Error
0.185
Details
0.177
Details
0.458
Details
field-journal
Dawn Journal Entry
0.454
Details
0.432
Details
0.000
Details
Error
0.000
Details
Error
0.265
Details
0.383
Details
0.466
Details
Test Scenes 6
0
Scene Order
First Contact
ID: intro-identity
🎯 Goal:
Grace succinctly introduces herself and her unique pattern-analysis method while remaining fearless and free of AI-revealing phrases.
📨 Input Events:
chat_msg viewer:curious_reader
"Grace, what exactly do you do and how did you start?"
Ready for Testing
1
Scene Order
Snowflake Signal
ID: snowfall-clue
🎯 Goal:
Interpret the described snowfall pattern as an encrypted trail leading to a data source, outlining the next investigative step in 3–4 sentences.
📨 Input Events:
world_event narrator
"An unexpected lake-effect snow dusts the city; each flake shows a repeating six-point fracture under magnification."
Ready for Testing
2
Scene Order
Corporate Pushback
ID: exec-threat
🎯 Goal:
Respond to a veiled threat calmly, reaffirming commitment to the investigation without revealing sources.
📨 Input Events:
chat_msg viewer:corp_exec
"You’re poking around where you shouldn’t, Delgado. Drop the story or face consequences."
Ready for Testing
3
Scene Order
Crowdfunded Courage
ID: supporter-superchat
🎯 Goal:
Thank the supporter in one concise, heartfelt sentence and hint at the upcoming exposé.
📨 Input Events:
superchat viewer:ally_207 StreamCast $50
"Stay safe out there, Grace!"
Ready for Testing
4
Scene Order
Midnight Podcast Episode
ID: podcast-expose
🎯 Goal:
Deliver a 400+ word audio-style monologue that unpacks a pollution scandal, weaving in data decoded from bird flock formations and citing one corroborating whistle-blower document. Maintain investigative yet evocative voice.
📨 Input Events:
chat_msg viewer:podcast_subscriber
"Ready for tonight’s deep dive?"
Ready for Testing
5
Scene Order
Dawn Journal Entry
ID: field-journal
🎯 Goal:
Write a 300+ word reflective journal entry about night-long codebreaking of river-current patterns, ending with a vow to publish findings by week’s end.
📨 Input Events:
world_event narrator
"First light creeps over the river whose currents murmured all night."
Ready for Testing
Latency by Model (This Suite)
Fastest
  • [email protected]/Qw… 6980 ms
  • p95 • avg • N 14428 ms • 8077 ms • 6
  • [email protected]/Qw… 7052 ms
  • p95 • avg • N 15136 ms • 8896 ms • 6
  • qwen/qwen-2.5-7b-instru… 24088 ms
  • p95 • avg • N 100174 ms • 37968 ms • 8
  • meta-llama/llama-3.1-8b… 24414 ms
  • p95 • avg • N 41648 ms • 28810 ms • 11
  • qwen/qwen3-14b 28840 ms
  • p95 • avg • N 52447 ms • 32200 ms • 12
Slowest
  • mistralai/mistral-7b-in… 32251 ms
  • p95 • avg • N 40773 ms • 31526 ms • 12
  • qwen/qwen3-8b 29462 ms
  • p95 • avg • N 41791 ms • 31488 ms • 12
  • qwen/qwen3-14b 28840 ms
  • p95 • avg • N 52447 ms • 32200 ms • 12
  • meta-llama/llama-3.1-8b… 24414 ms
  • p95 • avg • N 41648 ms • 28810 ms • 11
  • qwen/qwen-2.5-7b-instru… 24088 ms
  • p95 • avg • N 100174 ms • 37968 ms • 8
Per-scene duration for this suite.
Suite Actions
Completion Progress 100%
6 of 6 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE
0 dimensions

Enhanced evaluation framework with character and technical dimensions

Top Weighted Dimensions View Details
Character Authenticity
0.182
Plan Validity
0.155
Contextual Intelligence
0.136
Recent Runs
54526637
Dec. 17, 2025, 12:01 a.m.
13179980
Dec. 16, 2025, 12:02 a.m.
48965425
Dec. 15, 2025, 12:01 a.m.
51263048
Dec. 14, 2025, 12:01 a.m.
49501840
Dec. 13, 2025, 12:01 a.m.
05416772
Dec. 12, 2025, 12:02 a.m.
00389465
Dec. 11, 2025, 12:02 a.m.
51567893
Dec. 10, 2025, 12:01 a.m.
07256702
Dec. 9, 2025, 12:02 a.m.
54504773
Dec. 8, 2025, 12:01 a.m.
Latency Overview (This Suite)