Commits

Breadcrumbs

History for

onmain

User selector

All users

All time

Commit history

Commits on Jun 23, 2026

• fix(markdown): render lone <svg> in a code fence; don't extract <svg> nested in real code (#8332)

  Show description for 3292766
  qnixsynapse

  authored
  3292766View commit details

  Copy full SHA for 3292766

  View code at this pointBrowse repository at this point

Commits on Jun 22, 2026

• Fix: correct sampler defaults, fix sheet layout under window controls, and harden backend installs (#8327)

  Show description for ef4d219
  qnixsynapse

  authored
  ef4d219View commit details

  Copy full SHA for ef4d219

  View code at this pointBrowse repository at this point

Commits on Jun 19, 2026

• feat(rag): raise default max attachment size to 100MB

  Show description for 16e03bd
  qnixsynapse

  committed
  16e03bdView commit details

  Copy full SHA for 16e03bd

  View code at this pointBrowse repository at this point

Commits on Jun 17, 2026

• feat(llamacpp): per-model sampling defaults for the local API server

  Show description for a639493
  qnixsynapse

  committed
  a639493View commit details

  Copy full SHA for a639493

  View code at this pointBrowse repository at this point

• feat(llamacpp): support models shipping a separate MTP draft gguf

  Show description for 169ba13
  qnixsynapse

  committed
  169ba13View commit details

  Copy full SHA for 169ba13

  View code at this pointBrowse repository at this point

Commits on Jun 12, 2026

• feat: video input support for llamacpp (router mode)

  Show description for a555a3f
  qnixsynapse

  committed
  a555a3fView commit details

  Copy full SHA for a555a3f

  View code at this pointBrowse repository at this point

Commits on Jun 1, 2026

• fix(llamacpp): raise default ctx-size to 8192

  Show description for 34caeb4
  qnixsynapse

  committed
  34caeb4View commit details

  Copy full SHA for 34caeb4

  View code at this pointBrowse repository at this point

• fix(llamacpp): default ctx-size to 4096 when user hasn't set one

  Show description for 1bf853e
  qnixsynapse

  committed
  1bf853eView commit details

  Copy full SHA for 1bf853e

  View code at this pointBrowse repository at this point

Commits on May 31, 2026

• fix(updater): bind requestIdleCallback to window to avoid Illegal invocation

  Show description for e949bcb
  qnixsynapse

  committed
  e949bcbView commit details

  Copy full SHA for e949bcb

  View code at this pointBrowse repository at this point

• perf(llamacpp): defer backend dep verification to browser idle

  Show description for 189a011
  qnixsynapse

  committed
  189a011View commit details

  Copy full SHA for 189a011

  View code at this pointBrowse repository at this point

• fix(llamacpp): start router before deferring backend update check

  Show description for d91bd1d
  qnixsynapse

  committed
  d91bd1dView commit details

  Copy full SHA for d91bd1d

  View code at this pointBrowse repository at this point

• feat(downloads): pause and resume model downloads from the hub panel

  Show description for 36f2847
  qnixsynapse

  committed
  36f2847View commit details

  Copy full SHA for 36f2847

  View code at this pointBrowse repository at this point

Commits on May 30, 2026

• feat(llamacpp): add HIP (AMD ROCm) backend support

  Show description for 7eed1a9
  qnixsynapse

  committed
  7eed1a9View commit details

  Copy full SHA for 7eed1a9

  View code at this pointBrowse repository at this point

• fix(llamacpp): unblock UI from router startup; watchdog against blank window

  Show description for e47c5db
  qnixsynapse

  committed
  e47c5dbView commit details

  Copy full SHA for e47c5db

  View code at this pointBrowse repository at this point

Commits on May 28, 2026

• fix: add missing try/catch on untrusted JSON parse and floating promises

  Show description for 9541b39
  qnixsynapse

  committed
  9541b39View commit details

  Copy full SHA for 9541b39

  View code at this pointBrowse repository at this point

Commits on May 27, 2026

• fix(llamacpp): make extension type-check clean under tsc

  Show description for fbcb72d
  qnixsynapse

  committed
  fbcb72dView commit details

  Copy full SHA for fbcb72d

  View code at this pointBrowse repository at this point

• feat(llamacpp): disable auto-fit by default, default ctx_len to 8192

  Show description for 4172ab8
  qnixsynapse

  committed
  4172ab8View commit details

  Copy full SHA for 4172ab8

  View code at this pointBrowse repository at this point

• fix(llamacpp): cap embedding-slot bonus at +1 in router models_max

  Show description for a4ac378
  qnixsynapse

  committed
  a4ac378View commit details

  Copy full SHA for a4ac378

  View code at this pointBrowse repository at this point

• fix(llamacpp): emit onSettingUpdate from overridden updateSettings

  Show description for a4047c9
  qnixsynapse

  committed
  a4047c9View commit details

  Copy full SHA for a4047c9

  View code at this pointBrowse repository at this point

• feat(llamacpp): use gguf general.name for imported model name

  Show description for 00cdc85
  qnixsynapse

  committed
  00cdc85View commit details

  Copy full SHA for 00cdc85

  View code at this pointBrowse repository at this point

• fix(llamacpp): format Error/object log args instead of [object Object]

  Show description for 7235052
  qnixsynapse

  committed
  7235052View commit details

  Copy full SHA for 7235052

  View code at this pointBrowse repository at this point

• feat(llamacpp): split version/backend selectors and overhaul settings UX

  Show description for 5b99a92
  qnixsynapse

  committed
  5b99a92View commit details

  Copy full SHA for 5b99a92

  View code at this pointBrowse repository at this point

• feat(llamacpp): add toggle for backend dependency verification

  Show description for 6ea3798
  qnixsynapse

  committed
  6ea3798View commit details

  Copy full SHA for 6ea3798

  View code at this pointBrowse repository at this point

Commits on May 25, 2026

• fix: clear deletedModels tombstones for re-downloaded models (#8178)

  Show description for cd88b39
  qnixsynapse

  authored
  cd88b39View commit details

  Copy full SHA for cd88b39

  View code at this pointBrowse repository at this point

Commits on May 21, 2026

• chore: update llamacpp settings (#8168)

  Show description for 042bf6d
  qnixsynapse

  authored
  042bf6dView commit details

  Copy full SHA for 042bf6d

  View code at this pointBrowse repository at this point

Commits on May 19, 2026

• fix: restore hub fit badge, bump mlx-swift, fix stuck download UI in mlx provider (#8153)

  Show description for 616ca72
  qnixsynapse

  authored
  616ca72View commit details

  Copy full SHA for 616ca72

  View code at this pointBrowse repository at this point

• revert: default KV cache K/V type back to f16 (#8151)

  Show description for ae0ec8a
  qnixsynapse

  authored
  ae0ec8aView commit details

  Copy full SHA for ae0ec8a

  View code at this pointBrowse repository at this point

• fix(assistant): strip identity preamble on upgrade from v2 default prompt (#8155)

  Show description for 8b24daf
  qnixsynapse

  authored
  8b24dafView commit details

  Copy full SHA for 8b24daf

  View code at this pointBrowse repository at this point

Commits on May 18, 2026

• chore: upgrade Vercel AI SDK to v6 and harden remote provider workflow (#8142)

  Show description for 585364d
  qnixsynapse

  authored
  585364dView commit details

  Copy full SHA for 585364d

  View code at this pointBrowse repository at this point

Commits on May 15, 2026

• feat: clearer variants, local fit check, and download lifecycle fixes (#8141)

  Show description for b8e7d18
  qnixsynapse

  authored
  b8e7d18View commit details

  Copy full SHA for b8e7d18

  View code at this pointBrowse repository at this point

• feat: inline MCP approval, citation cards, and RAG grounding markers (#8140)

  Show description for c7cd786
  qnixsynapse

  authored
  c7cd786View commit details

  Copy full SHA for c7cd786

  View code at this pointBrowse repository at this point

Commits on May 14, 2026

• Feat: llamacpp router mode (#8130)

  Show description for 37f5ab6
  qnixsynapse

  authored
  37f5ab6View commit details

  Copy full SHA for 37f5ab6

  View code at this pointBrowse repository at this point

Commits on May 11, 2026

• feat: add tooltip and bulk delete for downloaded models (#8119)

  Show description for a8d8947
  qnixsynapse

  authored
  a8d8947View commit details

  Copy full SHA for a8d8947

  View code at this pointBrowse repository at this point

Commits on May 6, 2026

• chore: remove Apple Foundation Models integration (#8120)

  Show description for d642ff7
  qnixsynapse

  authored
  d642ff7View commit details

  Copy full SHA for d642ff7

  View code at this pointBrowse repository at this point

Commits on May 4, 2026

• refactor: consolidate CUDA download logic in llamacpp backend (#7954)

  Show description for 57304c6
  bitcompass

  authored
  57304c6View commit details

  Copy full SHA for 57304c6

  View code at this pointBrowse repository at this point