python C:\projects\Threadwell\scripts\twitter\tw.py graph   --id <snowflake>
python C:\projects\Threadwell\scripts\twitter\tw.py refresh --id <snowflake> --tip
python C:\projects\Threadwell\scripts\twitter\tw.py refresh --id <spine-tip> --tip --branch <branch-tip>
python C:\projects\Threadwell\scripts\twitter\tw.py add-branch --id <original-spine-tip> --from <reply-node>
python C:\projects\Threadwell\scripts\twitter\tw.py emit --id <spine-tip> --tip --preserve-existing
python C:\projects\Threadwell\scripts\twitter\tw.py audit-media --id <snowflake>
python C:\projects\Threadwell\scripts\twitter\tw.py publish --id <snowflake>
python C:\projects\Threadwell\scripts\twitter\tw.py sync    --handle <handle>
python C:\projects\Threadwell\scripts\twitter\tw.py backup  --id <snowflake>
python C:\projects\Threadwell\scripts\twitter\tw.py fallback --id <snowflake> --media-id <media-id> --role <role> --confirm-origin-unavailable
python C:\projects\Threadwell\scripts\twitter\tw.py restore-origin --id <snowflake> --media-id <media-id>
python C:\projects\Threadwell\scripts\twitter\tw.py migrate-media --id <snowflake>
python C:\projects\Threadwell\scripts\twitter\tw.py migrate-media --all --apply
python C:\projects\Threadwell\scripts\twitter\tw.py reslug --all
python C:\projects\Threadwell\scripts\twitter\tw.py reslug --all --apply

paths.py sets vault = two parents above this file, dumps = ../manual_slop/docs/twitter, scratch = ../Threadwell-ai/scratch.

refresh = refetch (cookies) + ingest + emit --force + media_merge. --tip treats --id as the tip and walks back to the OP regardless of handle. Repeatable --branch <tip-id> captures only those additional explicit tip paths and merges them into the same thread_data.json; it does not discover replies automatically. Captures run sequentially with gallery-dl retries disabled and five-second extractor/request pacing. Each thread emits exactly one archive directory owned by the OP (the post with reply_to_id == None); cross-author responders do not get a directory. Does not flip draft. Does not commit. Frozen ids abort.

emit --preserve-existing is the safe completion step after inspecting a fresh scratch capture. Fresh records win for ids the provider returned; posts present only in the existing emitted thread_data.json are retained after the command verifies both captures share a conversation root. This prevents an incomplete gallery-dl refresh from deleting archived replies. The merged scratch dataset keeps the requested --id as its stored spine tip before normal emission and media-manifest reconciliation.

add-branch incrementally extends an already emitted thread. --id must be the original stored spine tip; repeatable --from <reply-node> values are each captured once. The command locally retains each node's missing attachment path and entire visible descendant subtree, merges only previously absent posts into the existing thread_data.json, downloads media for the new posts, and re-emits with the original spine unchanged. It preserves existing origin/local/fallback/derived/OCR media records and adds provider-origin plus local-copy records for new branch media. A visible leaf is only the deepest reply returned by that capture; the command does not query descendants individually or claim provider-wide completeness. Gallery-dl may return only the ancestor chain when replies are behind nested show-more cursors; then the command adds nothing and known reply/stream-tip ids must be supplied explicitly. Fallback upload remains a separate explicit action after origin unavailability is confirmed.

Usual publish path after you like the note:

tw.py graph --id … — confirm tip vs root.
tw.py refresh --id … --tip if the dump stored a tip as root.
tw.py audit-media --id … — confirm origin URLs are selected.
tw.py publish --id … sets draft: false. You commit.

Site wikilinks must be [[archive/threads/<handle>/<date-slug>]]. Short [[slug]] 404s.

Media policy

Published Twitter/X media cites the original provider's HTTPS URL by default. The vault also retains a local, unpublished copy under assets/threads/<handle>/<date-slug>/ using the same relative taxonomy as the archive. A separately requested backup may mirror those assets to configured storage. Alternative-host uploads are a manual fallback used only after the origin is confirmed unavailable; the original URL and every fallback URL remain recorded in media.json.

The lift command is retired. Use fallback for a single media item after explicit confirmation, or restore-origin to select the immutable provider URL again. The Catbox userhash and backup destination id/root are read from secrets/credentials.toml by the script. Never paste them into chat.

Frozen threads (ids listed in do_not_refetch.txt, matched against every captured post id) are read-only across refetch, refresh, emit, manifest migration, note rewrite, fallback, publication selection, and tracked backup operations. Read-only audit-media may still inspect them.

Canonical thread directories

Each thread archive directory is named from the rendered title that appears in index.md (<YYYY-MM-DD>-<slugified rendered title>). tw.py reslug --all reads each existing index.md frontmatter date and title, derives the expected directory, and reports every mapping without writing. tw.py reslug --all --apply preflights every pair, refuses the entire apply on any non-frozen conflict, and moves assets first with the archive directory immediately after; if the archive move fails, assets roll back. If a destination is occupied, the apply aborts without -2, -3, or any other suffix. Frozen ids in do_not_refetch.txt are reported as frozen and skipped.

Future emits use the same rendered title for the directory and index.md frontmatter, so normal emit does this automatically — no flag is needed on emit, refresh, or add-branch. Hand-move only the archive or only the asset is not supported; both must move as a pair. Each old archive prefix is replaced exactly once in mutable *.md and *.canvas outside .git/, site/, assets/, secrets/, node_modules/, and frozen archive dirs, so mappings do not cascade. No aliases, redirect pages, compatibility folders, suffix fallback, or network contact.

Pieces

Script	Job
`tw.py`	locate / graph / refetch / add-branch / emit / refresh / ocr / merge / sync / publish / migrate-media / backup / fallback / restore-origin
`graph_dry_run.py`	tip graph, no vault writes
`ingest_gallery.py`	gallery-dl JSON → `thread_data.json`
`emit_archive.py`	JSON → notes + assets. Uses canonical media locations. `--force` preserves derived/fallback rows
`media_migrate.py`	legacy `media.json` → canonical locations, dry-run or apply
`media_audit.py`	local integrity, reference counts, mirror freshness, optional origin check
`media_manifest.py`	schema, validation, merge, selection, atomic I/O
`media_refs.py`	media-aware markup and note rewrite planning
`backup_assets.py`	sparse, hash-verified copy of one thread asset dir
`catbox_client.py`	one sanitized Catbox upload; no batching or retry
`fallback_media.py`	confirm-origin fallback activation and restore-origin
`frozen.py`	resolve frozen ids against every captured post id
`lift_catbox.py`	retired entry; points operators to `tw.py fallback`
`ocr_pass.py`	Umi HTTP `:1224` / tesseract / Windows OCR
`media_embed.py`	`--attach-ocr` only; refuses `--show crt`
`media_merge.py`	restore derived rows under canonical API
`crt_pass.py`	stock slang only; ShaderGlass is the look reference
`do_not_refetch.txt`	frozen id list
`tests/`	isolated, deterministic unit tests

Secrets: secrets/twitter_cookies.txt, secrets/credentials.toml [catbox] userhash and [backup] id/root. The script reads them; never paste into chat, tests, or commits.