Observed 2026-10-31. Partial trial results; live cross-provider acceptance is Not run.
An application signed-in flag does not identify the intended bot or prove API usability. Resolve the existing bot identity through the owned session before a send. Nuncio passed all 2205 tests and its full verification gate with mise exec bun@1.4.2. Its earlier HTTP2 fixture failure was resolved by correcting the child runtime; changing PATH alone was insufficient.
| Candidate | Version | Observed result |
|---|---|---|
| Nuncio-hq/grokbot-cli | 0.5.1 | 2205 tests passed plus lint, types, docs and builds using mise exec bun@1.4.2 |
| ScriptedAlchemy/grok-bot-cli | 0.12.4 | 442 unit and 97 route tests passed after a three-line history fixture isolation repair |
| aa22396584/trinity-cli | 0.4.6 | 197 offline tests executed: 190 passed, 7 optional real Muse tests skipped |
Pinned sources
- Nuncio-hq/grokbot-cli at 476ab6697d5d
- ScriptedAlchemy/grok-bot-cli at 43499fa27f36
- aa22396584/trinity-cli at 0609683213f7
Reproduce the tested scope
Nuncio: mise exec bun@1.4.2 -- bun run verify. ScriptedAlchemy: npm run test:unit and npm run test:routes after applying the reviewed history fixture isolation patch. Trinity: run its pytest suite in an isolated environment; optional real Muse contracts remain skipped. Grok identity reading was denied by automatic approval review before execution; action-specific confirmation is pending.
Evidence and reproduction
Generated from an actual ELF evidence import, receipt c1c1e681db8844553dc9f372. The source observation is a reviewed public projection of terminal tests and source reviews, not a recorded browser demonstration. Resolve the intended existing accounts before live trials.
Use the installed elf-communications-validation skill for request correlation and recovery rules. Run canonical agent-workflow validate on the owned store before regenerating. Published source and local generation do not establish live guide deployment.