Tool · MiMo Code · Xiaomi
read
How this differs from most implementations
- Staleness check before writeinvariantholdsvs fails in 6 of 8
- Output wrapper<path>/<type>/<content>vs plain in 6 of 8
Derived by comparing every recorded value against the other implementations of the same capability. Differing is not a defect unless the row is marked an invariant.
Observed · Read
compare all →| Output shape | |
| Dedicated read tool | readsrc |
| Line numbering | N: src |
| Output wrapper | <path>/<type>/<content>src |
| Limits | |
| Default line cap | 2000src |
| Byte / char cap | 50 KBsrc |
| Per-line cap | 2000 charssrc |
| Whichever hits first | lines or bytessrc |
| Pagination | |
| Offset base | 1-indexedsrc |
| Announces truncationinvariant | holdssrc (Showing X-Y of N. Use offset=Z) |
| Continuation hint | (Showing X-Y of N. Use offset=Z)src |
| Reports total lines | yessrc |
| Out-of-range offset | explicit error w/ line countsrc |
| Non-text | |
| Binary content never enters context silentlyinvariant | holdssrc Explicit error: Cannot read binary file. |
| Binary files | error: Cannot read binary filesrc |
| Images | attachment + mime sniffsrc |
| PDFs | attachmentsrc |
| Notebooks | (notebook-edit tool)src |
| Giant single line | per-line capsrc |
| Errors | |
| Missing path | Did you mean one of these?src |
| Path is a directory | lists entriessrc |
| Empty file | ? |
| Non-vision model | ? |
| Correctness | |
| Mixed line endings | ? |
| Staleness check before writeinvariant | holdssrc Tracks read state per file. |
| Side effects | |
| Warms LSP on read | yessrc |
| Injects nearby instruction files | AGENTS.md as system-remindersrc |