Repository navigation
chore(dashboard): trim pi, vibe, hermes, kimi and grok model pickers - #1161
Merged
Merged
Conversation
- pi: DeepSeek V4.1 Flash (new default) + V4 Pro; V4 Flash dropped. DeepSeek's own API now serves only V4.1 Flash (released 2026-09-10) and V4 Pro, and there is no V4.1 Pro. - vibe: Mistral Medium 3.5 (default, still Mistral's newest) + DeepSeek V4.1 Flash. - hermes: default, Claude Sonnet 5.5, GPT-6.1 Sol (replaces GPT-6 Sol, matching codex); Sonnet 4.6, GPT-5.4 and Gemini 2.5 Pro dropped. - Pi runtime default moved to deepseek/deepseek-v4.1-flash in resolve-harness.sh, the pi.sh meta block (harnesses.json regenerated), tests and docs. - Dropped ids stay workflow_dispatch choices and get RETIRED_MODEL_LABELS, so existing pins still dispatch and render as "(configured)".
- kimi: Kimi K2.7 Code (new default, same price as K2.6) + Kimi K3. K2.6 and K2.5 dropped (Moonshot's own API no longer lists K2.5). - grok: Grok 4.7 (default) + Grok 4.6; Grok 4.5 dropped. - Kimi runtime default moved to moonshotai/kimi-k2.7-code in resolve-harness.sh, the kimi.sh meta block (harnesses.json regenerated), docs. The Moonshot API-key path (install-harness.sh native-key) now pins Moonshot's native kimi-k2.7-code instead of kimi-k2.6. - Dropped ids stay workflow_dispatch choices with RETIRED_MODEL_LABELS. - Also fixes the pi row of the docs/harnesses.md credential table, missed in the previous commit.
aaronjmars
changed the title
Co-Authored-By: Claude <noreply@anthropic.com>
Scoped splice from an eyebrow v0.5.6 rescan (CI pin): only the .claude/skills/aeon artifact's file hashes and contentHash change; capabilities unchanged. The entry was already stale on main (SKILL.md, references/ci.md, references/skill-anatomy.md).
5 tasks done
aaronjmars
added a commit
that referenced
this pull request
Oct 3, 2026
* docs: sync PRs #1150-#1161 to aeon docs CHANGELOG [Unreleased] for #1150-#1161; drop the removed install-from-atrium from the skill-install.sh header; feature skill lists gpt-6.1-sol; dashboard README notes the phone layout and the Strategy/Soul reload after builder runs. * docs: proof banner says 8 community skill packs #1155 cut the listed community packs from 16 to 8, so the README proof-of-work banner and its alt text drop the stale "15+" claim. Only the number region of the image changed.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Trims the dashboard model pickers for pi, vibe (Mistral), hermes, kimi and grok.
What
deepseek/deepseek-v4.1-flash(new default) anddeepseek/deepseek-v4-pro. V4 Flash dropped.mistralai/mistral-medium-3-5(default) anddeepseek/deepseek-v4.1-flash(replaces V4 Flash).default,anthropic/claude-sonnet-5.5,openai/gpt-6.1-sol(replaces gpt-6-sol, matching codex). Sonnet 4.6, GPT-5.4 and Gemini 2.5 Pro dropped.moonshotai/kimi-k2.7-code(new default) andmoonshotai/kimi-k3. K2.6 and K2.5 dropped.grok-4.7(default) andgrok-4.6. Grok 4.5 dropped.resolve-harness.sh, thepi.sh/kimi.shmeta blocks (harnesses.jsonregenerated viabin/generate-harnesses-json), tests and docs.install-harness.shnative-key) now pins Moonshot's nativekimi-k2.7-codeinstead ofkimi-k2.6. The Moonshot login path keeps the account default, as before.workflow_dispatchmodel choices and getRETIRED_MODEL_LABELSentries, so existing pins still dispatch and render as "(configured)".Why
Checked 2026-10-03: DeepSeek's own API now serves only V4.1 Flash (
deepseek-flash, released 2026-09-10) and V4 Pro; there is no V4.1 Pro. Mistral Medium 3.5 is still Mistral's newest. Kimi K2.7 Code costs the same as K2.6 ($0.95 / $4 per 1M) and is on both OpenRouter and Moonshot's own API (native idkimi-k2.7-code); Moonshot's API no longer lists K2.5. Pickers now show only the latest models.Checks (local)
scripts/tests/test_resolve_harness.sh,test_dashboard_model_choices.sh,test_credential_manifest.sh,test_install_harness.sh,test_skill_health_routing.sh: passapps/dashboard:npm ci, typecheck, lint (0 errors, existing warnings only), test (300/300), build: pass