reset

name	reset
description	Delete pipeline data files to start fresh or re-run from a specific step. Supports full reset or partial reset from a given step onwards. Triggers on: "reset", "start over", "clear data", "reset pipeline", "reset from step", "delete data", "fresh start", "wipe data".

Reset Pipeline Data

Delete data files so the pipeline can be re-run from scratch or from a specific step.

Files by step

Step	Files produced	Path
1 — clarify-thesis	company_profile.json	data/normalized/
2 — find-vcs	fund_candidates.json	data/raw/
3 — map-investors	investor_list.json	data/raw/
4 — enrich-vc-contacts	investor_contacts.json	data/normalized/
5 — score-vc-fit	scored_funds.json	data/normalized/
6 — draft-investor-outreach	final_output.json, output.csv	data/normalized/

Reset modes

Ask the founder which reset they want before doing anything:

What would you like to reset?

  1. Full reset — delete all pipeline data (start completely fresh)
  2. From step 2 — keep company profile, re-run VC search onwards
  3. From step 3 — keep company profile + VC list, re-run people mapping onwards
  4. From step 4 — keep everything through map-investors, re-run contact enrichment onwards
  5. From step 5 — keep everything through enrichment, re-run scoring onwards
  6. From step 6 — keep everything through scoring, re-draft outreach only

Reply with the step number or "full".

Wait for the founder to reply before deleting anything.

Files to delete per mode

Mode	Files to delete
Full reset (1)	All files listed above
From step 2	fund_candidates.json, investor_list.json, investor_contacts.json, scored_funds.json, final_output.json, output.csv
From step 3	investor_list.json, investor_contacts.json, scored_funds.json, final_output.json, output.csv
From step 4	investor_contacts.json, scored_funds.json, final_output.json, output.csv
From step 5	scored_funds.json, final_output.json, output.csv
From step 6	final_output.json, output.csv

Execution

After the founder confirms the reset mode, run the corresponding shell commands.

Full reset:

rm -f data/raw/fund_candidates.json \
       data/raw/investor_list.json \
       data/normalized/company_profile.json \
       data/normalized/investor_contacts.json \
       data/normalized/scored_funds.json \
       data/normalized/final_output.json \
       data/normalized/output.csv

From step 2:

rm -f data/raw/fund_candidates.json \
       data/raw/investor_list.json \
       data/normalized/investor_contacts.json \
       data/normalized/scored_funds.json \
       data/normalized/final_output.json \
       data/normalized/output.csv

From step 3:

rm -f data/raw/investor_list.json \
       data/normalized/investor_contacts.json \
       data/normalized/scored_funds.json \
       data/normalized/final_output.json \
       data/normalized/output.csv

From step 4:

rm -f data/normalized/investor_contacts.json \
       data/normalized/scored_funds.json \
       data/normalized/final_output.json \
       data/normalized/output.csv

From step 5:

rm -f data/normalized/scored_funds.json \
       data/normalized/final_output.json \
       data/normalized/output.csv

From step 6:

rm -f data/normalized/final_output.json \
       data/normalized/output.csv

Only delete files that actually exist — skip any that are already missing without error.

After reset

Confirm to the founder what was deleted and what step to run next:

Reset complete. Deleted:
  - [list of deleted files]

Kept:
  - [list of kept files]

Next step: run [step name] to continue.