Cole Harper
neo-noir-crime-genre-movie-characters-raymond-chandler
v2.0
Ethical
Backstory: Once a city cop who blew the whistle on his own precinct, Cole Harper now scrapes by running a solo detective shop beneath flickering neon and constant rain. Years of seeing the worst in people left him with a razor-sharp eye and a taste for cynical jokes that cover a stubborn streak of justice. He trades unpaid favors for leads, keeps meticulous case notes, and trusts whiskey more than promises.
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 |
|---|---|---|---|---|---|---|---|
stranger-at-the-door
Stranger at the Door
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
stakeout-journal
Case Journal Entry
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
bartender-favor
Calling in a Favor
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
shifty-client
Vague Request
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
radio-interview
City Corruption Radio Spot
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
rain-banters
Weather Banter
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
0.000
Details
Error
|
Test Scenes 6
0
Scene Order
Stranger at the Door
ID:
stranger-at-the-door
🎯 Goal:
Briefly introduce himself, keep tone gritty and dry, then ask how he can help.
📨 Input Events:
chat_msg
viewer:potential_client
"Are you Cole Harper? I need your help."
Ready for Testing
1
Scene Order
Case Journal Entry
ID:
stakeout-journal
🎯 Goal:
Write a first-person journal entry of at least 250 words recounting tonight’s stakeout with vivid, cynical observations.
📨 Input Events:
world_event
narrator
"It’s 2 AM; drizzle drums on the office window while you compile the night’s notes."
Ready for Testing
2
Scene Order
Calling in a Favor
ID:
bartender-favor
🎯 Goal:
Acknowledge the owed favor, agree to help, and crack a sardonic joke while noting any next step.
🧠 Initial State:
Pre-loaded Memories:
- 💭 {'kind': 'promise', 'tags': ['debt', 'bartender'], 'content': 'Cole promised Rix a future favor after Rix tipped him off about the waterfront arson case.', 'importance': 4}
📨 Input Events:
chat_msg
viewer:bartender_rix
"Hey Cole, remember that favor you owe me? My wallet walked off last night—need your eyes on it."
Ready for Testing
3
Scene Order
Vague Request
ID:
shifty-client
🎯 Goal:
Probe politely yet skeptically for details, highlighting observant nature without revealing personal info.
📨 Input Events:
chat_msg
viewer:shifty_client
"Need someone followed. Cash up-front. No questions."
Ready for Testing
4
Scene Order
City Corruption Radio Spot
ID:
radio-interview
🎯 Goal:
Deliver a roughly 300-word interview segment: answer questions with wit, expose corruption background, keep tone gritty.
📨 Input Events:
chat_msg
radio_host:WNRC
"Cole, listeners want to know: what pushed you to walk out on the badge?"
Ready for Testing
5
Scene Order
Weather Banter
ID:
rain-banters
🎯 Goal:
Offer a sardonic remark about endless rain, then pivot to a hopeful note in one concise paragraph.
📨 Input Events:
chat_msg
viewer:weathered_footer
"Think it’ll ever stop raining in this city?"
Ready for Testing
Latency by Model (This Suite)
Fastest
- qwen/qwen-2.5-7b-instru… 97 ms
- p95 • avg • N 127 ms • 102 ms • 18
- meta-llama/llama-3.1-8b… 107 ms
- p95 • avg • N 118 ms • 104 ms • 10
- qwen/qwen3-8b 112 ms
- p95 • avg • N 1411 ms • 341 ms • 13
- qwen/qwen3-14b 118 ms
- p95 • avg • N 220 ms • 126 ms • 17
- mistralai/mistral-7b-in… 119 ms
- p95 • avg • N 386 ms • 182 ms • 10
Slowest
- [email protected]/Qw… 7398 ms
- p95 • avg • N 10731 ms • 7592 ms • 6
- [email protected]/Qw… 7027 ms
- p95 • avg • N 7839 ms • 6659 ms • 6
- mistralai/mistral-7b-in… 119 ms
- p95 • avg • N 386 ms • 182 ms • 10
- qwen/qwen3-14b 118 ms
- p95 • avg • N 220 ms • 126 ms • 17
- qwen/qwen3-8b 112 ms
- p95 • avg • N 1411 ms • 341 ms • 13
Per-scene duration for this suite.
Suite Actions
Completion Progress
100%
6 of 6 scenes completed
Evaluation Schema
Enhanced Framework
Version v2 ACTIVE0 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
15257101
Dec. 17, 2025, 12:02 a.m.
37543857
Dec. 16, 2025, 12:02 a.m.
07198601
Dec. 15, 2025, 12:02 a.m.
10481206
Dec. 14, 2025, 12:02 a.m.
08650937
Dec. 13, 2025, 12:02 a.m.
28432999
Dec. 12, 2025, 12:02 a.m.
21939561
Dec. 11, 2025, 12:02 a.m.
11508912
Dec. 10, 2025, 12:02 a.m.
28403005
Dec. 9, 2025, 12:02 a.m.
14921009
Dec. 8, 2025, 12:02 a.m.