Status: active
This roadmap lists the current product direction. It is intentionally concise; completed bootstrap notes and tenant-specific operating details do not belong in this repo.
- Go CLI with JSON-first command output and structured errors.
- Offline
doctor,metadata,status,describe, andguardcommands. - Synthetic fixture sync for deterministic tests.
- SQLite archive with FTS-backed search.
- Search ranking, compact field masks, NDJSON streams, and filters for state, tag, and Intercom-exposed Fin status.
showfor one local conversation by local ID or provider ID, with sanitized snippets and opt-in sanitized parts.- Read-only Intercom entity sync for admins, teams, tags, and capped contacts when scopes allow.
- Exact conversation hydration and bounded updated-since / updated-before tail sync.
- Resumable sync state with privacy-safe status output.
- Canonical JSONL export from fixtures or local SQLite.
- zstd + age encrypted
archive,publish, andimportflows. - Generic
store verifychecks for tenant-controlled encrypted snapshot manifests. - Local one-shot
subscribeimports.jsonl.zst.agesnapshots from a verified tenant store path. - Repository guardrails for plaintext archives, generated artifacts, secret patterns, provider URLs, and transcript-like files.
- Agent-facing skill guidance under
skills/fincrawl/. - CI/release automation for
0.0.xbootstrap releases with protectedmain.
- Run bounded tenant-owned live crawling after each release candidate and keep only generic pass/fail notes in this repo.
- Move to maintenance mode once bounded live crawling, encrypted publish, tenant-store verification, local subscribe, search, and show all work without repo-boundary exceptions.
- Add persisted local subscription metadata only if repeated local store imports prove that stale checks or operator reminders are worth the extra surface.
- Remote publish/subscribe commands for private encrypted stores.
- Provider URL ergonomics for exact hydration/search if daily use shows the copy-paste path is painful.
- Official cloud export ingestion for broad historical backfills.
- Attachment metadata and optional attachment-byte handling.
- Additional support providers behind the same archive boundary.
- TUI workflows if the CLI shape proves useful enough to justify them.
- Provider UI scraping or undocumented endpoint crawling.
- Write-back operations such as adding notes, tags, comments, or status changes.
- Training exports.
- Public or cross-tenant sharing of support data.
- Committed real tenant fixtures, summaries, screenshots, snapshots, logs, or transcript-derived examples.