Demand-gated backlog (no milestone)
- Dominant language
- Go
- Stars
- 218
- Forks
- 14
- PR merge metrics
- No merged PRs in 30d
Description
Items deliberately parked until user demand justifies them — revisit at each release:
- **In-grid editing** (rainfrog's top ask, harlequin PRs in flight — converging table stakes; needs Exec + PK detection + SQL-preview/txn UX) — includes #45
- **Sessions/workspaces** (per-project tabs/layout/history)
- **ClickHouse dialect** (no demand evidence yet)
- **NL->SQL chat panel** (agents own agentic SQL via `dbui mcp`; litecli/mycli's \llm moved nothing)
- **Native keyring lib** (dsn_cmd covers it)
- **Second AI backend / abstraction**
Contributor guide
Research direction
No files, tests, or implementation entry points are named. Start by reviewing the parked items and the referenced #45 context, then choose only one item after demand justifies it. Done would require a separately scoped feature with clear acceptance criteria; this issue itself is intentionally deferred.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- clickhouse, go
- Domain
- ai, cli, databases
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 20/100