name: threads-archiver
description: Archive and maintain Meta Threads post history into structured Markdown (index + per-post records) with stable URLs, dates, and content. Use when asked to scrape/export a Threads profile, backfill missing posts, deduplicate archives, convert raw post data to note format, or update a knowledge base (for example Obsidian) with chronological Threads records.

Threads Archiver

Archive Threads data in a repeatable, reviewable format.

Core Workflow

1. Identify archive scope.
2. Collect post data.
3. Normalize records.
4. Merge and deduplicate.
5. Render/update markdown output.
6. Validate ordering and coverage.

1) Identify archive scope

Capture these inputs before processing:

• Target account handle (for example @macau.drive.exam)
• Output note path
• Mode: full, incremental, or repair
• Required fields: date, url, content (minimum)

If scope is unclear, ask one short clarification question, then proceed.

2) Collect post data

Prefer one of these sources:

• Existing exported list (CSV/JSON/JSONL)
• Previously archived markdown note
• Browser-captured rows copied by the user

When using browser capture, gather at least:

• Canonical post URL
• Post date/time (or best available date)
• Full visible text content (or best-effort text)

3) Normalize records

Normalize each record into this logical schema:

• date: ISO date (YYYY-MM-DD) when possible
• url: canonical Threads URL
• content: post text (trimmed, preserve line breaks where meaningful)

Use scripts/build_threads_archive.py to normalize CSV/JSON/JSONL inputs and emit markdown-ready entries.

4) Merge and deduplicate

Deduplicate by URL first.

If URL is missing, use fallback key:

• date + first 80 chars of content

When duplicates conflict:

• Keep the record with richer content (longer non-whitespace text)
• Keep canonical URL variant

5) Render/update markdown output

Use the format in references/archive-format.md.

Default output sections:

• Metadata header (account + generated time)
• Chronological index (newest first)
• Detailed per-post entries (newest first)

For incremental updates:

• Insert only new posts
• Preserve existing manually edited commentary blocks if present

6) Validate ordering and coverage

Validate before finalizing:

• Dates sorted newest → oldest
• Every detailed post has a matching index row
• No duplicate URLs
• No empty content unless explicitly unavailable

Report summary:

• Total posts
• New posts added
• Duplicates removed
• Earliest and latest post date in archive

Resources

scripts/

• build_threads_archive.py: Convert raw post exports (CSV/JSON/JSONL) into normalized records and markdown sections.

references/

• archive-format.md: Canonical markdown template and formatting rules for index + detailed entries.