@thincoder/core 0.9.1

package/model-ref.mjs

@@ -0,0 +1,66 @@
+/**
+ * model-ref.mjs — composite model reference parser (MODEL-MERGE-SESSION; v2 2026-09-10).
+ * A model is an explicit "provider:model" composite — config.defaultModel (top level,
+ * new-session starting point) and session slots both speak this language. 有效域 = **provider 存在
+ * + 双段非空**：清单不再人工维护（模型清单由 provider 运行期 `GET /models` 拉取决定——
+ * PROVIDER.md §16），候选成员校验**已废除**——显式 `p:m` 一律放行（含多冒号首分割）。
+ * providers[].model 是渠道默认模型（单值）——候选清单字段 models[] 已整字段退场。
+ *
+ * Strict first-colon split — deliberately does NOT reuse resolveChildProvider's loose
+ * three-state resolution: ① legacy single-provider configs and ② model-name-only refs
+ * are gone with activeProvider/activeModel (F-1), so a bare provider or bare model is
+ * invalid everywhere (裁定③——裸 provider 拒——显式 p:m)。
+ *
+ * Runtime model resolution:
+ *   parseModelRef(ref, providers)      → { ok, provider, model } | { ok:false, reason }
+ *   resolveRuntimeProvider(providers, defaultModel)
+ *                                      → provider object with `.model` set, or {} when
+ *                                        defaultModel is null/invalid (D-S1: callers mark
+ *                                        _providerInvalid — never throws)
+ *   defaultModelReason(providers, defaultModel) → human reason for the D-S1 marker
+ *
+ * Consumers: config.mjs loadConfig (re-exported as the config hub), make-agent.mjs.
+ */
+export function parseModelRef(ref, providers) {
+  if (typeof ref !== "string" || !ref.trim()) {
+    return { ok: false, reason: 'model reference expected as "provider:model" (e.g. deepseek:deepseek-v4-pro)' }
+  }
+  // 首冒号分割——`a:b:c` → provider=a、model=`b:c`（`ollama:llama3:70b` 式模型名可用，M4）
+  const sep = ref.indexOf(":")
+  if (sep <= 0) {
+    return { ok: false, reason: `invalid model reference "${ref}" — expected provider:model (bare provider/model names are not accepted)` }
+  }
+  const providerName = ref.slice(0, sep)
+  const model = ref.slice(sep + 1)
+  if (!model) {
+    return { ok: false, reason: `invalid model reference "${ref}" — model part is empty (expected provider:model)` }
+  }
+  const list = Array.isArray(providers) ? providers : []
+  const provider = list.find((p) => p?.name === providerName)
+  if (!provider) {
+    const available = list.map((p) => p.name).join(", ") || "(none)"
+    return { ok: false, reason: `unknown provider "${providerName}" in model reference (available: ${available})` }
+  }
+  return { ok: true, provider, model }
+}
+
+/** Resolve the runtime provider object for config.defaultModel: provider clone carrying
+ *  `.model` = the parsed concrete model (API/spec consumers read provider.model unchanged).
+ *  Invalid/missing defaultModel → {} (D-S1 shape — name/model/baseURL absent so
+ *  validateProvider flags it; TUI/headless surface the reason from defaultModelReason). */
+export function resolveRuntimeProvider(providers, defaultModel) {
+  const r = parseModelRef(defaultModel, providers)
+  if (!r.ok) return {}
+  return { ...r.provider, model: r.model }
+}
+
+/** Human-readable reason behind an invalid/empty runtime resolution (D-S1 原因). */
+export function defaultModelReason(providers, defaultModel) {
+  const list = Array.isArray(providers) ? providers : []
+  if (!defaultModel) {
+    return list.length > 0
+      ? "defaultModel 未设置（config 顶层 defaultModel — 新会话起点；/config → 默认模型 设置）"
+      : "未配置任何 provider"
+  }
+  return parseModelRef(defaultModel, list).reason ?? `defaultModel "${defaultModel}" 无效`
+}

package/model-specs.mjs

@@ -0,0 +1,277 @@
+/**
+ * model-specs.mjs — known model capability table + spec lookup (2026-08-31 extract).
+ *
+ * Split from config.mjs (which had grown to 358 lines, past the 300 advisory
+ * line — TODO #1). config.mjs re-exports specForModel so the 23 existing
+ * importers stay untouched. (PROVIDER_PRESETS has since moved out, to
+ * config-presets.mjs — the core's single preset face, re-exported by config.mjs.)
+ */
+
+/**
+ * Known model capability spec table (prefix match, longer first).
+ * Used for compaction threshold derivation, continuation protocol selection, and capability-aware optimization.
+ *
+ * context:           context window (tokens)
+ * maxOutput:         max output tokens (defaults to context)
+ * thinking:          whether thinking/reasoning mode is supported
+ * partialMode:       Kimi/Qwen Partial Mode truncation continuation (assistant message with partial:true)
+ * prefixMode:        DeepSeek Prefix Completion truncation continuation (uses /beta endpoint, with prefix:true)
+ * multimodal:        whether multimodal (image/vision input supported)
+ * cacheMode:         context caching mode: "auto"=automatic / "prompt"=needs explicit / "none"=unsupported
+ * thinkApi:          thinking API type: "type"=thinking.type field / "effort"=reasoning_effort field
+ * thinkEnabledValue: when thinkApi is "type", the value used to enable thinking (default "enabled"; MiniMax uses "adaptive")
+ * reasoningEcho:     reasoning_content cross-turn echo strategy: "required"=must echo (error if missing) / "optional"=echo optional (default: don't echo)
+ * reasoningEffortEnum: valid reasoning_effort enum values (if undeclared, no validation — passed through as-is)
+ * tempRange:         valid temperature range [min, max] (if undeclared, no clamping)
+ */
+const MODEL_SPECS = [
+  // DeepSeek V4.1 series (2026-09-11). `deepseek-flash` = DeepSeek-V4.1-Flash (in service):
+  // 1M ctx / 384K out, thinking on by default (effort low/high/max), Chat Prefix Completion
+  // beta, automatic disk cache, vision. The two legacy names below are RETIRED — the server
+  // still accepts them and serves them from V4.1-Flash today (no switch window), so their
+  // rows carry the V4.1-Flash parameters (v4-flash gains multimodal).
+  ["deepseek-flash",    { context: 1_000_000, maxOutput: 384_000, thinking: true,  prefixMode: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 2], multimodal: true }],
+// deepseek-v4.1-flash = qwen-plan channel name for DeepSeek V4.1-Flash (token-plan GET /models verified 2026-09-15) — same fields as the row above; ".1" ≠ "-" so plain prefix lookup misses the existing rows (D-PR25)
+  ["deepseek-v4.1-flash", { context: 1_000_000, maxOutput: 384_000, thinking: true,  prefixMode: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 2], multimodal: true }],
+  // deepseek-v4-pro (V4-Pro-0813): fields unchanged — vision capability NOT verified, so no
+  // multimodal (conservative). From 2026-09-14 12:00 Beijing all requests route to V4.1-Flash.
+  ["deepseek-v4-pro",   { context: 1_000_000, maxOutput: 384_000, thinking: true,  prefixMode: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 2] }],
+  // deepseek-v4-flash: RETIRED name — still accepted, served by V4.1-Flash today → aligned row
+  ["deepseek-v4-flash", { context: 1_000_000, maxOutput: 384_000, thinking: true,  prefixMode: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 2], multimodal: true }],
+  // deepseek-v4-flash-vision-exp: retired experimental vision name — still accepted (V4.1-Flash)
+  ["deepseek-v4-flash-vision-exp", { context: 1_000_000, maxOutput: 384_000, thinking: true,  prefixMode: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 2], multimodal: true }],
+  // Kimi series
+  ["kimi-k3",           { context: 1_000_000, maxOutput: 131_072, thinking: true,  partialMode: true, multimodal: true, cacheMode: "auto",  thinkApi: "effort", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"] }],
+  // Qwen router prefixes model IDs with provider namespace: kimi/kimi-k3 → kimi-k3 (IK7K4V)
+  ["kimi/kimi-k3",      { context: 1_000_000, maxOutput: 131_072, thinking: true,  partialMode: true, multimodal: true, cacheMode: "auto",  thinkApi: "effort", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"] }],
+  // Kimi For Coding endpoint uses the short model ID "k3" (same specs as kimi-k3) — IK5VGJ
+  ["k3",                { context: 1_000_000, maxOutput: 131_072, thinking: true,  partialMode: true, multimodal: true, cacheMode: "auto",  thinkApi: "effort", reasoningEcho: "required", reasoningEffortEnum: ["low", "high", "max"] }],
+  // GLM series
+  // GLM-5.3: thinking always-on (no "disabled"); effort converges to low/high/max — NOT the
+  //          7-level glm-5.2 enum (verified vs docs.bigmodel.cn GLM-5.3 page, 2026-08)
+  ["glm-5.3",           { context: 1_000_000, maxOutput: 128_000, thinking: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 1], noUsageStream: true }],
+  ["glm-5.3-flash",     { context: 1_000_000, maxOutput: 128_000, thinking: true, multimodal: true, cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 1], noUsageStream: true }],
+  // glm-5.3-flashx（GLM-5.3-FlashX，2026-09 上线）——独立行（不再蹭 flash 前缀行）：maxOutput 131_072 =
+  // **校验级**（400「max_tokens…限制数值范围[1,131072]」）；effort 枚举 ["low","high","max"] = **校验级**
+  // （400 原文点名：low/high/max 受理，其余拒绝）。multimodal = **实测**（8×8 纯红 PNG → 答「红色」）。
+  // thinking = **实测**且**始终思考**（`thinking:{type:"disabled"}` → 400；裸请求默认开）——off 动作在
+  // UI 侧为 no-op（服务端无关闭路径）。其余 = **族沿用** flash 行（context 网络口径未探边；
+  // cacheMode / thinkApi / reasoningEcho / tempRange / noUsageStream = 前缀命中路径日常在用）。
+  // flash 行 128_000 系其自身口径，不得对齐本行（AC-3 零回归）。
+  ["glm-5.3-flashx",    { context: 1_000_000, maxOutput: 131_072, thinking: true, multimodal: true, cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", reasoningEffortEnum: ["low", "high", "max"], tempRange: [0, 1], noUsageStream: true }],
+  ["glm-5.2",           { context: 1_000_000, maxOutput: 128_000, thinking: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", reasoningEffortEnum: ["max", "xhigh", "high", "medium", "low", "minimal", "none"], tempRange: [0, 1], noUsageStream: true }],
+  ["glm-5",             { context: 1_000_000, maxOutput: 128_000, thinking: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", reasoningEffortEnum: ["max", "xhigh", "high", "medium", "low", "minimal", "none"], tempRange: [0, 1], noUsageStream: true }],
+  ["glm-4",             { context: 128_000,   maxOutput: 32_000,  thinking: true,  cacheMode: "auto", thinkApi: "type", reasoningEcho: "optional", tempRange: [0, 1], noUsageStream: true }],
+  // GPT series
+  ["gpt-5.6-sol",       { context: 1_050_000, maxOutput: 128_000, thinking: false, multimodal: true, cacheMode: "prompt" }],
+  ["gpt-5.6",           { context: 1_050_000, maxOutput: 128_000, thinking: false, multimodal: true, cacheMode: "prompt" }],
+  ["gpt-4.1",           { context: 1_000_000, maxOutput: 128_000, thinking: false, cacheMode: "prompt" }],
+  ["gpt-4o",            { context: 128_000,   maxOutput: 16_000,  thinking: false, multimodal: true, cacheMode: "prompt" }],
+  // Qwen series
+  // qwen3.7-max rejects image parts outright (DashScope 400 "Unexpected item type in content") — text-only
+  ["qwen3.7-max",       { context: 1_000_000, maxOutput: 131_072, thinking: true, partialMode: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["xhigh", "high"], tempRange: [0, 2] }],
+  // qwen3.7-flash / qwen3.8-flash (2026-09-20): thinking + effort enum + maxOutput + image intake
+  // are server-MEASURED (batch 2026-09-20-qwen-flash-specs §1.2 — bare request carries reasoning_tokens
+  // 166/27; 400 literal lists the enum; max_tokens range [1, 131072]). 3.7-flash enum has NO "max"
+  // (6 levels), 3.8-flash adds it (7); enum order = the server's literal order, first member "none"
+  // (D-5) — never re-sort by strength. context = doc/web statement (1M class, not API-measured);
+  // partialMode / cacheMode / thinkApi / tempRange = same-family carry-over (not independently
+  // measured on these rows); reasoningEcho stays undeclared (cross-turn echo unverified — R-4).
+  ["qwen3.7-flash",     { context: 1_000_000, maxOutput: 131_072, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  ["qwen3.8-flash",     { context: 1_000_000, maxOutput: 131_072, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"], tempRange: [0, 2] }],
+  ["qwen3.8-max",       { context: 1_000_000, maxOutput: 131_072, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["xhigh", "medium", "low"], tempRange: [0, 2] }],
+  // qwen3.8-omni-flash / qwen3.8-27b — parent 80-token probe (batch §1.8-①): reasoning_content
+  // present, reasoning_tokens 49, `effort:"none"` removes it ⇒ thinking MEASURED; enum / maxOutput
+  // / image intake measured the same way. omni-flash: context = doc/web statement (1M class, not
+  // API-measured); partialMode / cacheMode / thinkApi / tempRange = same-family carry-over;
+  // reasoningEcho undeclared (cross-turn echo unverified — R-4). The server also takes audio, this
+  // repo's call path sends images only ⇒ multimodal below = image intake (no `modalities` field, §2.7).
+  ["qwen3.8-omni-flash",{ context: 1_000_000, maxOutput: 131_072, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"], tempRange: [0, 2] }],
+  // 27b: context 262_144 = doc/web statement, the native value (1M is YaRN extrapolation, local
+  // deploy only); partialMode / cacheMode / thinkApi / tempRange same-family carry-over; reasoningEcho
+  // undeclared (cross-turn echo unverified — R-4).
+  ["qwen3.8-27b",       { context: 262_144,   maxOutput: 131_072, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"], tempRange: [0, 2] }],
+  // No generic "qwen" fallback row: unlisted qwen names deliberately resolve to DEFAULT_SPEC
+  // (128K / 32K, no vision) — per-name disposition in docs/core/design/MODEL-SPECS.md §2.4.
+  // Do not re-add a blanket prefix row (silent inheritance is how missing rows stayed invisible).
+
+  // qwen3.6 系五行（批 2026-09-20-qwen36-family-rows · 五名同形，取证叙述落此一处全量）：
+  // effort 枚举六档 = **校验级**（`max` → 400 原文点名 none/minimal/low/medium/high/xhigh——无 max，
+  // ≠ 3.8 系七档）；maxOutput 65_536 = **校验级**（400「Range of max_tokens should be [1, 65536]」；
+  // 65536 → 200）。thinking / multimodal = **实测**（裸请求 rc=143–239 默认开、none/minimal rc=0 可关；
+  // 64×64 纯红 PNG 五名受理）。context = **官方口径**（未探边，文档口径非 API 实测）；partialMode /
+  // cacheMode / thinkApi / tempRange = **族沿用**（qwen3.7/3.8 行先例）；reasoningEcho 不声明
+  // （跨轮回声未验——R-4）。日期戳快照名不登——前缀命中母名行（有意继承）。
+  ["qwen3.6-flash",     { context: 1_000_000, maxOutput: 65_536, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  // qwen3.6-plus：同形行（共享证据块见 flash 行上）——校验级 maxOutput / 六档枚举、官方口径 context、族沿用机制位。
+  ["qwen3.6-plus",      { context: 1_000_000, maxOutput: 65_536, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  // qwen3.6-max-preview：同形行——校验级 maxOutput / 六档枚举、官方口径 context、族沿用机制位；视觉通道实测 ✓ 但纯红图答「黑色」（抽样波动，如实记）。
+  ["qwen3.6-max-preview", { context: 1_000_000, maxOutput: 65_536, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  // qwen3.6-27b：同形行——校验级 maxOutput / 六档枚举、官方口径 context（未探边）、族沿用机制位。
+  ["qwen3.6-27b",        { context: 1_000_000, maxOutput: 65_536, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  // qwen3.6-35b-a3b：MoE 后缀（a3b = 激活参数量级）——校验级 maxOutput / 六档枚举、官方口径 context、族沿用机制位。
+  ["qwen3.6-35b-a3b",    { context: 1_000_000, maxOutput: 65_536, thinking: true, partialMode: true, multimodal: true, cacheMode: "none", thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh"], tempRange: [0, 2] }],
+  // MiniMax series
+  ["MiniMax-M3",        { context: 1_000_000, maxOutput: 128_000, thinking: true,  multimodal: true, cacheMode: "auto", thinkApi: "type", thinkEnabledValue: "adaptive", tempRange: [0, 2], noUsageStream: true }],
+  // MiMo series (Xiaomi — OpenAI-compatible https://api.xiaomimimo.com/v1;
+  // deep thinking via thinking.type, default ON; multi-turn tool calls MUST echo
+  // reasoning_content back exactly like DeepSeek V4, else 400 on follow-ups)
+  ["mimo-v2.5-pro",     { context: 1_000_000, maxOutput: 128_000, thinking: true,  thinkApi: "type", reasoningEcho: "required", tempRange: [0, 1.5] }],
+  ["mimo-v2.5",         { context: 1_000_000, maxOutput: 128_000, thinking: true,  multimodal: true, thinkApi: "type", reasoningEcho: "required", tempRange: [0, 1.5] }],
+  ["minimax-m3",        { context: 1_000_000, maxOutput: 128_000, thinking: true,  multimodal: true, cacheMode: "auto", thinkApi: "type", thinkEnabledValue: "adaptive", tempRange: [0, 2], noUsageStream: true }],
+  ["minimax-m1",        { context: 256_000,   maxOutput: 128_000, thinking: false, cacheMode: "auto", noUsageStream: true }],
+  // Grok series (xAI — OpenAI-compatible)
+  // grok-4.x: 500K context per xAI Grok 4.6 spec (corrected 2026-08; earlier entries said 1M)
+  ["grok-4.6",          { context: 500_000,   maxOutput: 64_000,  thinking: false, multimodal: true, tempRange: [0, 2] }],
+  ["grok-4.5",          { context: 500_000,   maxOutput: 64_000,  thinking: false, multimodal: true, tempRange: [0, 2] }],
+  ["grok-4",            { context: 500_000,   maxOutput: 64_000,  thinking: false, multimodal: true, tempRange: [0, 2] }],
+  ["grok-4-mini",       { context: 128_000,   maxOutput: 16_000,  thinking: false, tempRange: [0, 2] }],
+  // Mistral series (OpenAI-compatible)
+  ["mistral-large",     { context: 128_000,   maxOutput: 32_000,  thinking: false, multimodal: true, tempRange: [0, 2] }],
+  ["codestral",         { context: 256_000,   maxOutput: 32_000,  thinking: false, tempRange: [0, 2] }],
+  // Claude series (Anthropic)
+  ["claude-opus-5",     { context: 1_000_000, maxOutput: 128_000, thinking: false, multimodal: true, cacheMode: "none", format: "anthropic" }],
+  ["claude-sonnet-5",   { context: 1_000_000, maxOutput: 128_000, thinking: false, multimodal: true, cacheMode: "none", format: "anthropic" }],
+  ["claude-opus-4",     { context: 200_000,   maxOutput: 32_000,  thinking: false, multimodal: true, cacheMode: "none", format: "anthropic" }],
+  ["claude-sonnet-4",   { context: 200_000,   maxOutput: 32_000,  thinking: false, multimodal: true, cacheMode: "none", format: "anthropic" }],
+  ["claude-3.5-haiku",  { context: 200_000,   maxOutput: 8_192,   thinking: false, cacheMode: "none", format: "anthropic" }],
+  // Gemini series (Google)
+  ["gemini-3-pro",      { context: 1_000_000, maxOutput: 64_000,  thinking: false, multimodal: true, cacheMode: "none", format: "google", noUsageStream: true }],
+  ["gemini-2.5-pro",    { context: 2_000_000, maxOutput: 64_000,  thinking: false, multimodal: true, cacheMode: "none", format: "google", noUsageStream: true }],
+  ["gemini-2.5-flash",  { context: 1_000_000, maxOutput: 64_000,  thinking: false, multimodal: true, cacheMode: "none", format: "google", noUsageStream: true }],
+  // Tencent Hunyuan (TokenHub — 聚合网关 `https://tokenhub.tencentmaas.com/v1`；批 2026-09-20-channel-onboarding)
+  // hy3：thinking / thinkApi / effort 枚举 = **本渠道实测**——裸请求 reasoning_content 在场（tok=16），
+  // `reasoning_effort:"none"` 即消失（实测唯一有效 off 路径）。枚举 = 七值**受理级探针**（七值全 200 受理；
+  // 乱值 400 泛化拒收、服务端不列枚举——等级低于 seed 的「服务端真校验」，两族同值集但等级不混）；
+  // 序 = 服务端原文序、首项 "none"，**禁按强度重排**（D-5）。multimodal：**不声明**——真值 32×32 纯红图
+  // 答 "Unknown"（对照组明说不识图）⇒ 实测无视觉。尺寸位两级逐条标（AC-2）：maxOutput 128_000 =
+  // **参考实配**（网关对 max_tokens 不硬拒 ⇒ 上限未证）；context 256_000 = **网络口径**（他仓/文档 256K，未本渠道实测）。
+  ["hy3",               { context: 256_000, maxOutput: 128_000, thinking: true, thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"] }],
+  // hy3-preview（在场实测：/models + 401 前验活；能力位**未探针**）——**尺寸行**：仅登记尺寸位，取值 =
+  // hy3 **同族沿用**（maxOutput 参考实配 / context 网络口径）。D-11：能力位不跨名沿用 ⇒ thinking / 枚举 / 视觉全不声明。
+  ["hy3-preview",       { context: 256_000, maxOutput: 128_000 }],
+  // hy4-preview（在场实测：/models + 401；能力位**未探针**）——尺寸行：取值 = hy3 族**同族沿用**
+  // （maxOutput 参考实配级取值 / context 网络口径；批次档 §1.2 未给该名尺寸行 ⇒ 上报清单 2，行注即标级）。
+  ["hy4-preview",       { context: 256_000, maxOutput: 128_000 }],
+  // Doubao Seed (Volcano Ark — `https://ark.cn-beijing.volces.com/api/v3`；批 2026-09-20-channel-onboarding)
+  // 两档全针实测：裸请求 reasoning_content 在场（code tok=219 / lite tok=161）、`effort:"none"` 即消失 ⇒
+  // thinking / thinkApi = 实测；枚举 = **服务端真校验级**（乱值 400 明列 invalid + 七档全 200）⇒ 七值即服务端值域；
+  // maxOutput 131_072 = **实测**（262 144 → 400 "above maximum"）；context 256_000 = **官方口径**（官方 256K）
+  // + **实测**下界 210K 输入受理（探至账号 429 停手）；multimodal = 实测（纯红图答 "Red"）。
+  ["doubao-seed-2-0-code-preview-260215", { context: 256_000, maxOutput: 131_072, thinking: true, thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"], multimodal: true }],
+  // seed-lite = 同族全针（批次档 §1.2 lite 行：思考 tok=161 / 视觉 "Red" / 上限 131 072 / echo 真名）——
+  // context 同按 **官方口径** + **实测**下界标注；行独立（同值集 ≠ 同行：改一行不动另一行）。
+  ["doubao-seed-2-0-lite-260428", { context: 256_000, maxOutput: 131_072, thinking: true, thinkApi: "effort", reasoningEffortEnum: ["none", "minimal", "low", "medium", "high", "xhigh", "max"], multimodal: true }],
+]
+const DEFAULT_SPEC = { context: 128_000, maxOutput: 32_000, cacheMode: "none" }
+
+/** Look up spec by model name prefix (case-insensitive), conservative default for unknown models.
+ *
+ * Vendor-namespace prefix stripping (2026-09-04)：第三方 token 市场（roapi/new-api/one-api/
+ * aiproxy 聚合网关）惯例在模型名前加厂商前缀（zhipu/glm-5.3、openai/gpt-4o）。完整名未命中
+ * 且含 "/" 时，剥掉第一个 "/" 前的 namespace 再按前缀匹配一次——ZHIPU/GLM-5.3 → glm-5.3 命中
+ * 真实规格，不再降级 128K 默认。kimi/kimi-k3 的显式 alias 行保留为文档锚（发送路径
+ * provider.core isRouter 依赖含 "/" 判定），通用机制已覆盖同类。 */
+const warnedModels = new Set() // warn once per model name — spec lookup runs on every request (hot path)
+// Pre-sorted once at module scope — spec lookup runs on every request (agent, provider core,
+// context, auto-think, TUI rendering); re-sorting per call was wasteful.
+const SORTED_SPECS = [...MODEL_SPECS].sort((a, b) => b[0].length - a[0].length)
+
+/** Single table lookup shared by specForModel / specMatch — prefix match (case-insensitive)
+ *  with vendor-namespace stripping. Returns null on a miss (DEFAULT_SPEC is the caller-side
+ *  fallback and is deliberately NOT returned here — `matched` needs the miss itself). */
+function lookupSpec(model) {
+  const m = (model ?? "").toLowerCase()
+  for (const [prefix, spec] of SORTED_SPECS) {
+    if (m.startsWith(prefix.toLowerCase())) return spec
+  }
+  // Vendor-namespace strip: vendor/model — retry the prefix match on the bare model part.
+  const slash = m.indexOf("/")
+  if (slash > 0) {
+    const bare = m.slice(slash + 1)
+    for (const [prefix, spec] of SORTED_SPECS) {
+      if (bare.startsWith(prefix.toLowerCase())) return spec
+    }
+  }
+  return null
+}
+
+/** Unknown model: warn ONCE (not per request) so a typo'd ID or a missing alias surfaces
+ *  instead of silently degrading to the 128K default (IK5VGJ). The dedupe set is shared by
+ *  specForModel / specMatch (PROVIDER.md §16 M5). */
+function warnUnknownModel(model) {
+  const m = (model ?? "").toLowerCase()
+  if (m && !warnedModels.has(m)) {
+    warnedModels.add(m)
+    console.warn(`[config] model "${model}" not found in MODEL_SPECS — using default spec (128K context, 32K output). Check the model ID or add an alias.`)
+  }
+}
+
+export function specForModel(model) {
+  const hit = lookupSpec(model)
+  if (hit) return hit
+  warnUnknownModel(model)
+  return DEFAULT_SPEC
+}
+
+/** specMatch(model) → { spec, matched }（PROVIDER.md §16 M5——切换回显的来源判定）。
+ *  `matched:false` = DEFAULT_SPEC 兜底（未知模型）；与 specForModel 共享同一查表实现（单次），
+ *  specForModel 的返回形状与共享对象契约零改（热路径零变）。 */
+export function specMatch(model) {
+  const hit = lookupSpec(model)
+  if (hit) return { spec: hit, matched: true }
+  warnUnknownModel(model)
+  return { spec: DEFAULT_SPEC, matched: false }
+}
+
+/**
+ * providerSpec(provider) — spec with a provider-level context override (PROVIDER.md §15, 2026-09-02).
+ *
+ * providers[].context is configured in K units (128 = 128K = 131072 tokens) and overrides the
+ * MODEL_SPECS value for THIS provider only — the same model can have different real context
+ * windows on different endpoints (official vs local deployment). The ×1024 conversion happens
+ * HERE and nowhere else.
+ *
+ * Returns a COPY ({ ...spec, context }) — the shared spec object from the SORTED_SPECS lookup
+ * must never be mutated, or the override would leak across providers (T-C1).
+ *
+ * Validation is defensive (pure function): absent/invalid context falls back to the plain spec
+ * (config.mjs loadConfig already warns + strips invalid values; this guard covers direct callers
+ * and keeps the function total). specForModel stays a pure table lookup — callers without a
+ * provider keep using it.
+ */
+export function providerSpec(provider) {
+  const spec = specForModel(provider?.model ?? "")
+  const k = Number(provider?.context) // Number() 接受数字字符串（"128"）——两端语义统一（code review #1）
+  if (Number.isInteger(k) && k > 0) return { ...spec, context: k * 1024 }
+  return spec
+}
+
+/**
+ * assistantToolCallMessage(response, spec) — the tool-round assistant message pushed onto the
+ * machine line (D-CC22, #109 — live-push face of the echo-safety family; single construction
+ * point, CONTEXT-COMPACTION.md §6.10 #9).
+ *
+ * `reasoningEcho:"required"` families (deepseek / kimi / mimo) MUST carry the reasoning echo on
+ * every tool-call assistant message: the field is NEVER omitted — an absent/non-string
+ * `response.reasoning` is echoed as `""` (live-shape probe 2026-09-20: missing field → 400
+ * "must be passed back"; empty string → 200 with reasoning still returned). `optional` /
+ * undeclared families (incl. DEFAULT_SPEC) never get the field (behavior byte-identical).
+ *
+ * Callers: thincoder-core/agent.mjs (main loop) · thincoder-core/advisor/loop.mjs (review
+ * mirror) · thincoder-vscode/src/agent.mjs (shell loop) — each passes its own spec face.
+ */
+export function assistantToolCallMessage(response, spec) {
+  const msg = {
+    role: "assistant",
+    content: response.content || null,
+    tool_calls: response.toolCalls.map((tc) => ({
+      id: tc.id, type: "function",
+      function: { name: tc.name, arguments: tc.arguments },
+    })),
+  }
+  if (spec?.reasoningEcho === "required") {
+    msg.reasoning_content = typeof response.reasoning === "string" ? response.reasoning : ""
+  }
+  return msg
+}

package/package.json

@@ -0,0 +1,32 @@
+{
+  "name": "@thincoder/core",
+  "version": "0.9.1",
+  "description": "ThinCoder shared core — shared mechanism modules + shared prompt content (CLI ↔ VS Code extension).",
+  "type": "module",
+  "scripts": {
+    "test": "node test/run.mjs",
+    "prepublishOnly": "npm test"
+  },
+  "license": "MIT",
+  "engines": {
+    "node": ">=22.13.0"
+  },
+  "exports": {
+    "./*": "./*"
+  },
+  "files": [
+    "*.mjs",
+    "advisor/",
+    "agent/",
+    "agent-tools/",
+    "git/",
+    "mcp/",
+    "memory/",
+    "provider/",
+    "session-*.mjs",
+    "tools/",
+    "traces/",
+    "prompts/",
+    "tool-docs/"
+  ]
+}