# Rename Suggestions for Conversation Files — APPLIED 2026-05-27

> **Status**: **APPLIED on 2026-05-27.** All four conversation files have been renamed to the schema below. This document now functions as a historical mapping from the original export names to the current filenames, useful if you ever need to match a file back to a Claude/Perplexity chat URL or to a v0/v1 archive document that still references the originals.

---

## Why rename at all?

Three reasons:

1. **Date legibility.** None of the current filenames expose the dates of the underlying conversation. To know whether a file documents v1 or v4 work, you have to open it. A `YYYY-MM-DD_` prefix fixes that at the directory listing level.
2. **Sort order.** Date-prefixed names sort chronologically by default, which mirrors the development arc.
3. **Provider clarity.** "Claude" vs "Perplexity" matters methodologically — the Perplexity sessions used `claude46sonnetthinking` in COPILOT mode (so the *underlying model* was also Claude, but the *interface and search-augmented behavior* were Perplexity's). Keeping the provider in the filename preserves that distinction.

## Why *not* rename?

The current filenames are the names Claude/Perplexity gave when exporting. If you ever need to match an exported file back to a specific chat URL or session, the original name is the natural key. Two mitigations are possible: (a) keep the original names and let the folder README map them to dates; or (b) rename and store the original-name mapping in this file. Option (b) is what is proposed below.

---

## Applied rename mapping

| # | Original export filename | Current filename | Provider | Model (where exposed) | Conversation date range |
|---|--------------------------|------------------|----------|----------------------|--------------------------|
| 1 | `perplexity_search- v1 and v2 banco_pt_process_tracing e v2 banco_pt_process_tracing_1779835784644.md` | `2026-05-02_perplexity_pt-literature-and-v1-v2-schema.md` | Perplexity | `claude46sonnetthinking` (COPILOT) + some `pplx_pro_upgraded` and `turbo` | 2026-05-02 23:09 → 2026-05-05 12:02 (11 questions) |
| 2 | `Claude-Banco de dados para process tracing com testes bayesianos.md` | `2026-05-03_claude_bayesian-tests_v1-through-v4.md` | Claude (claude.ai) | not exposed in export header | 2026-05-03 11:16 → 2026-05-24 14:03 (exported 2026-05-26 17:08) |
| 3 | `perplexity_search-v3 Banco Process Tracing_1779831341921.md` | `2026-05-05_perplexity_v3-structural-pivot.md` | Perplexity | `claude46sonnetthinking` (COPILOT) | 2026-05-05 13:18 → 2026-05-24 22:32 (7 questions) |
| 4 | `Claude-V5 Banco de dados para process tracing em pesquisa.md` | `2026-05-24_claude_v5-fairfield-charman-2022-deepread.md` | Claude (claude.ai) | not exposed in export header | 2026-05-24 20:08 → 2026-05-26 10:45 |

### Filename schema

```
YYYY-MM-DD_{provider}_{short-descriptor}.md
```

- **Date**: ISO 8601 of the *first* message in the conversation.
- **Provider**: `claude` or `perplexity` (lowercase, no underscore).
- **Short descriptor**: hyphen-separated, lowercase, ≤ 6 words, indicates what the conversation *was about* rather than its sequence number (descriptors like "v1-v2" or "v3-pivot" are acceptable).

This is intentionally close to the convention proposed in plan v2 of `golden-scribbling-lake.md`, but adjusted after seeing the actual content:

- Conversation #1 covers both literature search **and** the v1/v2 schema (not just literature search), so the descriptor reflects both.
- Conversation #2 actually carried the database from v1 through v4 inside a single chat (not just "bayesian-tests"), so the descriptor is more explicit.
- Conversation #4's distinctive feature is the upload of the F&C 2022 book and the *deep* operationalization of it — captured as `fairfield-charman-2022-deepread`.

---

## What the renaming would *not* do

- Not rename `xlsx_templates/` — already renamed during the reorganization.
- Not touch the original Claude/Perplexity export timestamps embedded *inside* each file (those are the source of truth).
- Not change the contents of any conversation file.

---

## Rename commands (executed 2026-05-27)

For the record, these are the exact `mv` commands that were run:

```bash
cd "llm_conversations_database-development/conversations"

mv "perplexity_search- v1 and v2 banco_pt_process_tracing e v2 banco_pt_process_tracing_1779835784644.md" \
   "2026-05-02_perplexity_pt-literature-and-v1-v2-schema.md"

mv "Claude-Banco de dados para process tracing com testes bayesianos.md" \
   "2026-05-03_claude_bayesian-tests_v1-through-v4.md"

mv "perplexity_search-v3 Banco Process Tracing_1779831341921.md" \
   "2026-05-05_perplexity_v3-structural-pivot.md"

mv "Claude-V5 Banco de dados para process tracing em pesquisa.md" \
   "2026-05-24_claude_v5-fairfield-charman-2022-deepread.md"
```

After running, the following files were patched to use the new names:

- `../README.md` (folder README)
- `../../README.md` (root README)
- `../../HISTORY.md`
- `../../NEWS.md`

The archived snapshots in `../../md_documentation_previous_versions/` were *not* touched — they are point-in-time records and continue to reference the original filenames (which is correct: those references describe the state of the world at the time the snapshot was taken).

---

## Open question — applies to the entire folder

The current folder name `llm_conversations_database-development` is descriptive but inconsistent with the otherwise short/lowercase convention used in `versions/` (now `md_documentation_previous_versions/`) and `conversations/`. If you want a shorter alternative:

- `dev-history/` — succinct, communicates "development history."
- `design-archive/` — emphasizes the archival nature.

Neither is necessary; the current name is workable. Listed here only for completeness.