sample-output-pack.md

Sample Output Pack

The sample output pack shows how Thesis OS feels in operation without exposing private data.

Path:

examples/sample_outputs/

All files in the pack are synthetic and marked:

sample: true
public_sanitized: true
not_financial_advice: true
source_policy: synthetic_example

Included Outputs

Output	Why It Exists
Thesis card	Shows how a living investment thesis is structured
Nightly Top 5 deep dive	Shows the collection -> screening -> judgment handoff
Concentrated strategy review	Shows portfolio-level common-driver thinking
Screener discovery results	Shows how Alpha compresses candidates
Screener performance feedback	Shows the closed loop from signal to outcome
Prediction ledger accountability	Shows hits and misses together, with process and result scores separated
Social collection summary	Shows how qualitative signals are summarized safely

Design Principle

Public examples should demonstrate the loop, not reveal the deployment.

The examples should never include:

• real account data
• private holdings or exact portfolio weights
• raw community messages
• private channel IDs or user handles
• secrets, tokens, cookies, OAuth state, or API keys
• paid-feed raw data
• non-public company information

How To Use

New contributors can read the sample outputs in this order:

1. screener-discovery-results.md
2. nightly-top5-deep-dive.md
3. thesis-card-ai-infrastructure-basket.md
4. nightly-concentration-strategy.md
5. screener-performance-feedback.md
6. prediction-ledger-accountability.md
7. social-collection-summary.md

This order follows the operating loop:

collect -> screen -> compress -> judge -> record -> evaluate -> improve