Torjus Håkestad
1755364bba
New MCP server that queries live Prometheus and Alertmanager HTTP APIs with 8 tools: list_alerts, get_alert, search_metrics, get_metric_metadata, query (PromQL), list_targets, list_silences, and create_silence. Extends the MCP core with ModeCustom and NewGenericServer for servers that don't require a database. Includes CLI with direct commands (alerts, query, targets, metrics), NixOS module, and comprehensive httptest-based tests. Bumps existing binaries to 0.2.1 due to shared internal/mcp change. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
26 lines
1.2 KiB
Markdown
26 lines
1.2 KiB
Markdown
# TODO - Future Improvements
|
|
|
|
## Usability
|
|
|
|
- [ ] Progress reporting during indexing ("Fetching nixpkgs... Parsing options... Indexing files...")
|
|
- [ ] Add `search_files` MCP tool - search for files by path pattern (e.g., find all nginx-related modules)
|
|
- [ ] Include file size metadata in `get_option` declarations (byte size and/or line count) so clients know file sizes before fetching
|
|
- [ ] Add range parameters to `get_file` (`offset`, `limit`) with sensible defaults (~200-300 lines) to avoid dumping massive files
|
|
|
|
## Robustness
|
|
|
|
- [ ] PostgreSQL integration tests with testcontainers (currently skipped without manual DB setup)
|
|
- [ ] Graceful handling of concurrent indexing (what happens if two clients index the same revision?)
|
|
|
|
## New MCP Servers
|
|
|
|
- [x] `nixpkgs-packages` - Index and search nixpkgs packages (implemented in `nixpkgs-search packages`)
|
|
- [x] `lab-monitoring` - Query Prometheus and Alertmanager APIs (8 tools, no database required)
|
|
|
|
## Nice to Have
|
|
|
|
- [ ] Option history/diff - compare options between two revisions ("what changed in services.nginx between 24.05 and 24.11?")
|
|
- [ ] Auto-cleanup - prune old revisions after N days or keep only N most recent
|
|
- [ ] Man page generation
|
|
- [ ] Shell completions (bash, zsh, fish)
|