Trace Companies

Step 2: Review script results

Check the summary output:

• How many repos were resolved by Layer 1 (known mapping)?
• How many by Layer 2 (GitHub API)?
• How many remain unknown?

If all repos are resolved (unknown = 0), skip to Step 4.

Step 3: LLM Web Search for remaining unknowns

For repos still marked unknown, use Web Search to determine company affiliation. Do NOT guess.

Batch optimization

• Group unknown repos by GitHub org — repos in the same org belong to the same company
• Research the org once, apply to all its repos

For each unknown org/repo:

1. Web search for "{org_name}" company or "{project_name}" developed by
2. Determine: which company (if any) owns/maintains this?
3. Record:
   • company_name — company name, or unknown if truly independent
   • evidence — actual URLs and facts from search
   • confidence — S/A/B/C

Decision criteria

• Has parent company: owned/maintained by a for-profit company → record it
• Independent project: community-driven, no single corporate owner → unknown
• University labs: not companies → unknown

Step 4: Present to user

Show results grouped by company for review. The user may correct assignments.

After user confirmation, update output/company.csv.

Step 5: Output summary

Print:

• Total repos processed
• Repos resolved by script (Layer 1 + Layer 2) vs LLM (Layer 3)
• Company distribution (count by company)
• Count by confidence level