can1357 520f6f9bcc feat(catalog): standardized reasoning effort configuration ...

- Transitioned DeepSeek models to an explicit `[High, Max]` effort ladder, removing stale alias maps.
- Updated the OpenAI compatibility layer to enforce authoritative `supportsReasoningEffort` and `omitReasoningEffort` flags.
- Synchronized `models.json` definitions to reflect accurate reasoning effort capabilities across the catalog.

2026-07-10 13:59:54 +02:00

..

advisor

fix(advisor): used uuidv7 codex session ids

2026-07-10 07:24:36 +00:00

async

fix(agent): bounded async job disposal

2026-06-30 23:37:11 +00:00

auto-thinking

fix(coding-agent): size title/commit/speech/classifier budgets for backends that ignore disableReasoning

2026-07-03 00:47:20 +00:00

autolearn

fix(agent): kept autolearn memory prompts cache-stable

2026-06-28 18:10:57 +00:00

autoresearch

fix(autoresearch): guarded before_agent_start against undefined event.systemPrompt

2026-06-27 15:51:31 +00:00

capability

Merge upstream/main into feat/profiles-and-alias

2026-06-11 13:07:52 -03:00

cli

fix(coding-agent): preserved fork prompt cache affinity

2026-07-10 07:22:28 +00:00

collab

fix collab reconnect backpressure drain

2026-07-05 13:25:26 +02:00

commands

chore: update changelogs

2026-07-06 04:17:21 +02:00

commit

fix(commit): capture split-executor staged diff with binary contents

2026-07-05 21:32:40 +00:00

config

feat: introduced max reasoning effort tier across model and rpc systems

2026-07-10 13:39:42 +02:00

dap

fix(dap): removed per-write exit handlers

2026-07-02 23:51:18 +02:00

debug

fix(debug): guarded v8.setFlagsFromString call so CPU profiler works on Bun

2026-06-30 12:47:07 +00:00

discovery

fix(discovery): preserve marketplace-root skill selection

2026-07-08 15:19:38 +02:00

edit

fix(coding-agent/edit): sealed inverse video and preserved gutters in wrapped diff rows

2026-07-06 00:34:23 +09:00

eval

fix(coding-agent): hardened same-realm runtime guards from PR review

2026-07-10 12:33:53 +02:00

exa

fix: address easy issues review

2026-06-08 15:47:57 -03:00

exec

fix(bash): avoided aborting native timeout signal

2026-07-10 03:48:40 +00:00

export

perf(export): skip TTSR delta buffering when no rule matches text/thinking

2026-07-02 17:29:50 +09:00

extensibility

fix(coding-agent): armed ask timeout for legacy ui

2026-07-10 04:13:18 +00:00

goals

refactor(coding-agent): streamlined codebase by deduplicating helper logic and shims

2026-07-02 02:40:08 +02:00

hindsight

fix(mnemopi): stripped retention markers from embeddings

2026-07-03 04:59:02 +00:00

internal-urls

fix(coding-agent): resolved agent:// for nested subagent output

2026-07-07 13:28:59 -03:00

irc

fix(session): auto-reply awaited IRC sends to an idle plan-mode session

2026-07-02 10:30:51 +02:00

jsonrpc

refactor(coding-agent): consolidated message framing logic

2026-06-22 05:29:00 +02:00

lib

refactor(catalog)!: split model catalog from pi-ai

2026-06-10 04:06:57 +02:00

lsp

fix(lsp): handled go.work workspace diagnostics

2026-07-10 07:14:49 +00:00

markit

fix(coding-agent): handled inline image pdf delimiters

2026-07-04 12:18:06 +00:00

mcp

fix(mcp): kept macos stdio servers attached

2026-07-09 21:39:42 +00:00

memories

feat(coding-agent): removed canonical model variant selection and tracking

2026-07-01 05:22:42 +02:00

memory-backend

fix(agent): kept autolearn memory prompts cache-stable

2026-06-28 18:10:57 +00:00

mnemopi

fix(mnemopi): made recall fact ids resolvable via memory reads

2026-07-09 18:27:23 +02:00

modes

feat(catalog): standardized reasoning effort configuration

2026-07-10 13:59:54 +02:00

plan-mode

fix(coding-agent): prevented asynchronous write races in plan reviews

2026-07-02 02:19:26 +02:00

prompts

feat(coding-agent): renamed explore agent to scout

2026-07-10 12:51:50 +02:00

registry

feat(coding-agent): supported advisor transcript persistence

2026-06-19 03:50:00 +02:00

secrets

fix(coding-agent): hid secrets from advisor prompts

2026-06-22 07:05:23 +00:00

session

feat: introduced max reasoning effort tier across model and rpc systems

2026-07-10 13:39:42 +02:00

slash-commands

Merge PR #4387: fix(fast): enable custom OpenAI-compatible providers (@roboomp)

2026-07-05 13:10:27 +02:00

ssh

Merge remote-tracking branch 'origin/farm/16f7ef23/ssh-helper-timeouts'

2026-07-02 23:43:11 +02:00

stt

style: bun run fix

2026-07-01 00:07:30 +00:00

subprocess

fix(tiny): repaired cuda sidecars for auto gpu requests

2026-07-03 18:34:56 +00:00

task

feat(coding-agent): renamed explore agent to scout

2026-07-10 12:51:50 +02:00

tiny

fix(tiny): repaired cuda side-runtime install

2026-07-03 18:21:19 +00:00

tool-discovery

refactor(coding-agent): migrated tool schema helpers to toolWireSchema

2026-06-18 01:57:23 +02:00

tools

Merge PR #4999: fix(coding-agent): restore ask tool timeout fallback

2026-07-10 12:37:36 +02:00

tts

feat(coding-agent): supported streaming TTS for long text inputs

2026-07-06 04:12:39 +02:00

tui

feat(coding-agent): refactored eval tool to single-step execution

2026-06-23 00:59:58 +02:00

utils

feat(coding-agent): renamed explore agent to scout

2026-07-10 12:51:50 +02:00

web

test(coding-agent): covered docs.rs gunzip cap via extracted seam

2026-07-05 13:31:22 +02:00

cli-commands.ts

feat: add manual coding-agent gc command

2026-06-26 08:39:50 +01:00

cli.ts

feat(coding-agent/tts): redesigned streaming speech vocalization

2026-07-02 08:30:58 +02:00

config.ts

…

cursor.ts

feat(coding-agent/tools)!: replaced paths arrays with path strings

2026-07-02 08:30:33 +02:00

index.ts

refactor(coding-agent): integrated zod validation and upgraded typebox shim

2026-06-18 01:57:23 +02:00

main.ts

fix(coding-agent): guarded fork cache on scoped models

2026-07-10 07:41:59 +00:00

priority.json

config(coding-agent): extended model priority lists

2026-06-27 04:34:39 +02:00

sdk.ts

feat: introduced max reasoning effort tier across model and rpc systems

2026-07-10 13:39:42 +02:00

startup-splash.ts

fix(coding-agent): tighten startup splash review coverage

2026-06-17 10:53:08 -04:00

system-prompt.test.ts

feat: introduced max reasoning effort tier across model and rpc systems

2026-07-10 13:39:42 +02:00

system-prompt.ts

merge PR #4644: fix(prompting): hide eval guidance when disabled

2026-07-08 15:19:36 +02:00

telemetry-export.ts

perf(coding-agent): deferred heavy module imports to startup paths

2026-06-06 23:25:18 +02:00

thinking.ts

feat: introduced max reasoning effort tier across model and rpc systems

2026-07-10 13:39:42 +02:00

workspace-tree.ts

feat: standardized elision markers and improve transcript viewer robustness

2026-06-19 04:35:18 +02:00