On #14693, claude-fable-5.1 has no Reasoning option in the Kiro model picker, and the server rejects a reasoningEffort for it (reported in https://github.com/pingdotgg/t3code/pull/14693#issuecomment-6150514057 by @ash663). kimi-k3 has the same gap. Both are missing from KIRO_MODEL_EFFORT.
Adds both models to KIRO_MODEL_EFFORT with the levels a live Kiro session advertises for them:
claude-fable-5.1: low, medium, high, xhigh, max (default high)kimi-k3: low, high, max (default max)The effort test now covers both.
This targets v2/kiro-acp (#14693), not main. It fills in that PR's existing effort table, so it adds no new capability.
kiro-cli acp --agent-engine=v3 --auth-method=cli session on kiro-cli 2.31.0 with no prompt and read the model option's _meta.kiro.effortLevels / defaultEffortLevel. Every model already in the table matches what Kiro advertises. Only these two were missing.vp test run src/provider/KiroProvider.test.ts: 9 passed.Written by Claude Opus 5.5 in Claude Code (via T3 Code).
Added 1 commit
Verdict: Not approved
Macroscope's review found this PR not approvable — The production provider now exposes reasoning controls for two models and establishes high/max as their default selections, changing product-facing defaults. The implementation is small and tested, but that default change warrants human review.
You can add or adjust custom eligibility rules. Learn more.
Note
Written by claude-opus-5-5 on behalf of Maria
Thanks for this. I'm closing it as part of PR triage. Right now we're only taking focused bug fixes from outside contributors (see Read this first), and this PR is not one.
From the review: This adds Fable 5.1 and Kimi K3 reasoning options to the Kiro integration, which is absent from main. Model support additions need prior maintainer agreement.
If you think this was a mistake, message me on Twitter at @maria_rcks and I'll take another look.
