jeo-code 0.6.26 → 0.6.27

package/CHANGELOG.md

@@ -6,6 +6,18 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 The README mirrors the latest 5 entries — regenerate with `bun run changelog:sync`.
 
+## [0.6.27] - 2026-06-19
+_Ponytail pass on the reasoning-tier mapper, plus a real-tmux verification of `jeo --tmux`._
+
+### Changed
+- **`thinkingToReasoningEffort` collapsed to its essential mapping (ponytail/YAGNI pass).** The four redundant pass-through branches (`minimal`/`low`/`medium`/`high` each returning themselves) are now a single `level === "xhigh" ? "high" : level` — behavior-identical (every level still maps to a genuine reasoning effort; only an unset level stays off), 8 fewer lines, fully covered by the existing `model-manager`/`round-b` contract tests. Reasoning continues to activate at EVERY thinking level (gajae parity).
+
+### Fixed
+- **`/agents <role> provider <name>` now accepts every registered provider and always shows a model list (jeo team role config).** Three compounding bugs surfaced via a real `jeo --tmux` session pinning a role to `groq`: (1) `isProviderName` was an unsound type guard hardcoding only 5 names (`anthropic|openai|gemini|antigravity|ollama`), so `/agents <role> provider groq` (and every other OpenAI-compat provider — deepseek, openrouter, mistral, …) was rejected as invalid usage; it now validates against the canonical `PROVIDER_NAMES` registry. (2) Live discovery only returns ids for a logged-in, reachable provider, and the catalog backfill applied only to OAuth-source providers — so an unconfigured API-key provider showed an EMPTY model list and silently pinned a bare default. (3) The 24 OpenAI-compat providers carry no capability-catalog rows, so even the catalog fallback was empty for them. The new `providerPickEntries` helper now climbs live ids → static catalog → the provider's known default model, so the list is never empty, and the source is labeled (`Live`/`Catalog … log in to list live models`). Verified end-to-end in a real tmux session (`#1 groq/llama-3.3-70b-versatile` listed and pinned). Covered by `test/provider-pick-entries.test.ts` and a new `isProviderName` regression test in `test/launch-flags.test.ts`.
+
+### Verified
+- **`jeo --tmux` session profile confirmed against the real `tmux` binary.** The gjc-parity profile (`mouse on`, `@jeo-profile`/`@jeo-branch`/`@jeo-project` markers, `set-clipboard on`, copy-mode `mode-style`) was exercised on an isolated `-L` socket using the exact `=name:` target syntax the launch code emits — every option set and read back correctly. `test/tmux.test.ts` passes 12/0 alongside the full 1645/0 suite.
+
 ## [0.6.26] - 2026-06-19
 _The forge emblem is redrawn again as the mascot crayfish, foregrounding its signature pincer claws (집게)._
 

package/README.ja.md

@@ -204,11 +204,11 @@ CI は `.github/workflows/npm-publish.yml` で公開します — GitHub リリ
 ## 変更履歴 (Changelog)
 
 <!-- CHANGELOG:START (auto-generated from CHANGELOG.md — run `bun run changelog:sync`) -->
+- **[0.6.27]** (2026-06-19) — Ponytail pass on the reasoning-tier mapper, plus a real-tmux verification of `jeo --tmux`.
 - **[0.6.26]** (2026-06-19) — The forge emblem is redrawn again as the mascot crayfish, foregrounding its signature pincer claws (집게).
 - **[0.6.25]** (2026-06-19) — Reasoning works at every thinking level (gajae parity), and the forge emblem is redrawn as the neon-lens coding wizard.
 - **[0.6.24]** (2026-06-19) — `/provider` opens an interactive onboarding selector (OAuth vs API-compatible), and OpenAI-compatible backends gain per-vendor native-reasoning formats.
 - **[0.6.23]** (2026-06-19) — Live reasoning/thinking streams in the TUI across every provider, three new OpenAI-compatible backends (LM Studio, xAI, Kimi) join the auth/discovery/catalog surface, and Gemini gains native function-calling.
-- **[0.6.22]** (2026-06-18) — Extended-thinking activation is now consistent across providers: a `low` session thinking level enables reasoning everywhere.
 
 See [CHANGELOG.md](CHANGELOG.md) for the full history.
 <!-- CHANGELOG:END -->

package/README.ko.md

@@ -204,11 +204,11 @@ CI는 `.github/workflows/npm-publish.yml`로 배포합니다 — GitHub 릴리
 ## 변경 이력 (Changelog)
 
 <!-- CHANGELOG:START (auto-generated from CHANGELOG.md — run `bun run changelog:sync`) -->
+- **[0.6.27]** (2026-06-19) — Ponytail pass on the reasoning-tier mapper, plus a real-tmux verification of `jeo --tmux`.
 - **[0.6.26]** (2026-06-19) — The forge emblem is redrawn again as the mascot crayfish, foregrounding its signature pincer claws (집게).
 - **[0.6.25]** (2026-06-19) — Reasoning works at every thinking level (gajae parity), and the forge emblem is redrawn as the neon-lens coding wizard.
 - **[0.6.24]** (2026-06-19) — `/provider` opens an interactive onboarding selector (OAuth vs API-compatible), and OpenAI-compatible backends gain per-vendor native-reasoning formats.
 - **[0.6.23]** (2026-06-19) — Live reasoning/thinking streams in the TUI across every provider, three new OpenAI-compatible backends (LM Studio, xAI, Kimi) join the auth/discovery/catalog surface, and Gemini gains native function-calling.
-- **[0.6.22]** (2026-06-18) — Extended-thinking activation is now consistent across providers: a `low` session thinking level enables reasoning everywhere.
 
 See [CHANGELOG.md](CHANGELOG.md) for the full history.
 <!-- CHANGELOG:END -->

package/README.md

@@ -204,11 +204,11 @@ Required npm token permissions (repository secret `NPM_TOKEN`):
 ## Changelog
 
 <!-- CHANGELOG:START (auto-generated from CHANGELOG.md — run `bun run changelog:sync`) -->
+- **[0.6.27]** (2026-06-19) — Ponytail pass on the reasoning-tier mapper, plus a real-tmux verification of `jeo --tmux`.
 - **[0.6.26]** (2026-06-19) — The forge emblem is redrawn again as the mascot crayfish, foregrounding its signature pincer claws (집게).
 - **[0.6.25]** (2026-06-19) — Reasoning works at every thinking level (gajae parity), and the forge emblem is redrawn as the neon-lens coding wizard.
 - **[0.6.24]** (2026-06-19) — `/provider` opens an interactive onboarding selector (OAuth vs API-compatible), and OpenAI-compatible backends gain per-vendor native-reasoning formats.
 - **[0.6.23]** (2026-06-19) — Live reasoning/thinking streams in the TUI across every provider, three new OpenAI-compatible backends (LM Studio, xAI, Kimi) join the auth/discovery/catalog surface, and Gemini gains native function-calling.
-- **[0.6.22]** (2026-06-18) — Extended-thinking activation is now consistent across providers: a `low` session thinking level enables reasoning everywhere.
 
 See [CHANGELOG.md](CHANGELOG.md) for the full history.
 <!-- CHANGELOG:END -->

package/README.zh.md

@@ -204,11 +204,11 @@ CI 通过 `.github/workflows/npm-publish.yml` 发布 — GitHub 发布 release
 ## 更新日志 (Changelog)
 
 <!-- CHANGELOG:START (auto-generated from CHANGELOG.md — run `bun run changelog:sync`) -->
+- **[0.6.27]** (2026-06-19) — Ponytail pass on the reasoning-tier mapper, plus a real-tmux verification of `jeo --tmux`.
 - **[0.6.26]** (2026-06-19) — The forge emblem is redrawn again as the mascot crayfish, foregrounding its signature pincer claws (집게).
 - **[0.6.25]** (2026-06-19) — Reasoning works at every thinking level (gajae parity), and the forge emblem is redrawn as the neon-lens coding wizard.
 - **[0.6.24]** (2026-06-19) — `/provider` opens an interactive onboarding selector (OAuth vs API-compatible), and OpenAI-compatible backends gain per-vendor native-reasoning formats.
 - **[0.6.23]** (2026-06-19) — Live reasoning/thinking streams in the TUI across every provider, three new OpenAI-compatible backends (LM Studio, xAI, Kimi) join the auth/discovery/catalog surface, and Gemini gains native function-calling.
-- **[0.6.22]** (2026-06-18) — Extended-thinking activation is now consistent across providers: a `low` session thinking level enables reasoning everywhere.
 
 See [CHANGELOG.md](CHANGELOG.md) for the full history.
 <!-- CHANGELOG:END -->

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "jeo-code",
-  "version": "0.6.26",
+  "version": "0.6.27",
   "description": "Clean, highly optimized AI coding agent using spec-first loop",
   "type": "module",
   "main": "src/cli.ts",

package/src/ai/model-manager.ts

@@ -100,18 +100,15 @@ export function thinkingMaxTokens(level?: "minimal" | "low" | "medium" | "high"
   return 16000;
 }
 
-/** Map the thinking level to an OpenAI reasoning-effort tier. `minimal` is preserved as a
- *  genuine (lightest) reasoning effort — NOT collapsed to `low` — so reasoning works at EVERY
- *  thinking level (gajae parity: Minimal is a real effort). Only an unset level returns undefined
- *  (reasoning off). `xhigh` maps to `high`, the deepest tier the provider APIs accept. */
+/** Map the thinking level to an OpenAI reasoning-effort tier. minimal/low/medium/high pass
+ *  through unchanged and xhigh folds to high (the deepest tier the provider APIs accept), so
+ *  reasoning works at EVERY thinking level (gajae parity: minimal is a real effort). Only an
+ *  unset level returns undefined (reasoning off — the explicit /fast path). */
 export function thinkingToReasoningEffort(
   level?: "minimal" | "low" | "medium" | "high" | "xhigh",
 ): "minimal" | "low" | "medium" | "high" | undefined {
   if (!level) return undefined;
-  if (level === "minimal") return "minimal";
-  if (level === "low") return "low";
-  if (level === "high" || level === "xhigh") return "high";
-  return "medium";
+  return level === "xhigh" ? "high" : level;
 }
 
 /** Describe a model id: alias expansion + the provider it routes to. For `/model` + diagnostics.

package/src/commands/launch/flags.ts

@@ -1,6 +1,6 @@
 import * as fs from "node:fs";
 import * as path from "node:path";
-import { type ProviderName, type ModelRole, type ThinkLevel, catalogMetadata } from "../../ai";
+import { type ProviderName, type ModelRole, type ThinkLevel, catalogMetadata, PROVIDER_NAMES } from "../../ai";
 
 export interface LaunchFlags {
   list: boolean;
@@ -39,7 +39,10 @@ function takeValue(args: string[], index: number, inlinePrefix: string): { value
 }
 
 export function isProviderName(input: string | undefined): input is ProviderName {
-  return input === "anthropic" || input === "openai" || input === "gemini" || input === "antigravity" || input === "ollama";
+  // Validate against the canonical registry, not a hand-maintained subset — the
+  // old 5-name list silently rejected every OpenAI-compat provider (groq,
+  // deepseek, openrouter, …) at `/agents <role> provider <name>`.
+  return input !== undefined && (PROVIDER_NAMES as readonly string[]).includes(input);
 }
 
 export function isThinkingLevel(input: string | undefined): input is ThinkLevel {

package/src/commands/launch.ts

@@ -36,7 +36,7 @@ import { callLlm, type Message } from "../agent/loop";
 import { friendlyProviderError } from "../util/provider-error";
 import { readGlobalConfig, saveConfigPatch } from "../agent/state";
 import { rememberModelPatch, recentModelsForDisplay } from "../agent/model-recency";
-import { describeModel, describeAllProviders, thinkingMaxTokens, thinkingToReasoningEffort, discoverModels, flattenModels, resolveSelection, catalogMetadata, resolveRoleModel, CODEX_MODELS, qualifyModelId } from "../ai";
+import { describeModel, describeAllProviders, thinkingMaxTokens, thinkingToReasoningEffort, discoverModels, flattenModels, resolveSelection, catalogMetadata, catalogByProvider, resolveRoleModel, CODEX_MODELS, qualifyModelId } from "../ai";
 import type { ProviderModelsResult, PickEntry, ProviderName, ModelRole, ThinkLevel } from "../ai";
 import { readGoalState, writeGoalState, clearGoalState, verifyGoal } from "../agent/goal-verifier";
 
@@ -232,6 +232,29 @@ function providerDefaultModel(p: ProviderName): string {
   return openaiCompatDef(p)?.defaultModel ?? STATIC_PROVIDER_DEFAULT[p] ?? "";
 }
 
+/**
+ * Pick-list entries for ONE provider, with static fallbacks so the list is never
+ * empty.
+ *
+ * Live discovery yields ids only for a logged-in, reachable provider, and
+ * `catalogOr` backfills the static catalog ONLY for OAuth sources — so an
+ * API-key/keyless provider that isn't configured yet had an empty list and was
+ * silently pinned to a bare default. Prefer live ids; else the provider's
+ * capability catalog; else its single known default model (all 24 OpenAI-compat
+ * providers carry one) so the user always sees at least one pickable id.
+ */
+export function providerPickEntries(live: ProviderModelsResult[], want: ProviderName): PickEntry[] {
+  const fromLive = flattenModels(live.filter(r => r.provider === want));
+  if (fromLive.length) return fromLive;
+  const catalog = catalogByProvider(want);
+  if (catalog.length) {
+    return catalog.map((m, i) => ({ index: i + 1, provider: want, model: qualifyModelId(m.providerModel, want) }));
+  }
+  const fallback = providerDefaultModel(want);
+  return fallback ? [{ index: 1, provider: want, model: qualifyModelId(fallback, want) }] : [];
+}
+
+
 
 export function formatResumeHint(sessionId: string): string {
   return `Resume with: jeo launch --resume ${sessionId}`;
@@ -3462,7 +3485,7 @@ export async function runLaunchCommand(args: string[]): Promise<void> {
         if (modelArg?.toLowerCase() === "provider") {
           const want = (tokens[2] ?? "").toLowerCase();
           if (!isProviderName(want)) {
-            console.log(`Usage: /agents ${role.id} provider <anthropic|openai|gemini|antigravity|ollama> [model|#N]`);
+            console.log(`Usage: /agents ${role.id} provider <name> [model|#N] — e.g. anthropic, openai, gemini, groq, deepseek, openrouter (any configured provider)`);
             continue;
           }
           const st = (await describeAllProviders()).find(s => s.name === want);
@@ -3475,7 +3498,8 @@ export async function runLaunchCommand(args: string[]): Promise<void> {
             }
           }
           const live = await getLiveModels();
-          const forProvider = flattenModels(live.filter(r => r.provider === want));
+          const forProvider = providerPickEntries(live, want);
+          const liveForProvider = live.some(r => r.ok && r.provider === want && r.models.length > 0);
           const explicit = tokens[3];
           let chosenModel: string;
           if (explicit && forProvider.length) {
@@ -3494,7 +3518,7 @@ export async function runLaunchCommand(args: string[]): Promise<void> {
           } else if (explicit) {
             chosenModel = qualifyModelId(explicit, want);
           } else if (forProvider.length) {
-            // No model given → the provider's first live model, provider-qualified.
+            // No model given → the provider's first known model, provider-qualified.
             chosenModel = qualifyModelId(forProvider[0]!.model, want);
           } else {
             chosenModel = providerDefaultModel(want);
@@ -3503,7 +3527,9 @@ export async function runLaunchCommand(args: string[]): Promise<void> {
           console.log(`${role.title} pinned to ${want} via model ${chosenModel} — saved to ~/.jeo/config.json`);
           if (forProvider.length) {
             lastPickIndex = forProvider;
-            console.log(`Live ${want} models — refine with /agents ${role.id} #N:`);
+            const sourceNote = liveForProvider ? "Live" : "Catalog";
+            const tail = liveForProvider ? "" : " (log in to list live models)";
+            console.log(`${sourceNote} ${want} models — refine with /agents ${role.id} #N:${tail}`);
             for (const line of formatPickListWithCapabilities(lastPickIndex, { current: chosenModel, cap: 12 })) console.log(line);
           }
           continue;

package/src/tui/components/ascii-art.ts

@@ -412,52 +412,48 @@ export async function animateFrames(stage: AsciiStage, opts: AnimateFramesOption
   }
   return total;
 }
-/** The compact jeo forge mark: a clean, wordless pictograph of the mascot — the
- *  neon crayfish (가재) from assets/character.png, read for its signature feature:
- *  the two raised pincer CLAWS (집게). Read top→bottom: the open pincer jaws
- *  (◣◣ / ◢◢) reaching up, the claw arms (◆══╲ ╱══◆) angling into the body where
- *  the blue ◆ knuckles meet, the head with its glowing eye/terminal cluster
- *  (◉◉◉), then the rounded carapace/tail (╲▔▔╱). The mark is purely symbolic —
- *  NO embedded lettering (the brand wordmark lives in the welcome header, not
- *  under the emblem); the JEO identity is carried by the crayfish-claw silhouette
- *  alone. Width-1 glyphs only (box drawing + geometrics) so padding/centering
- *  math stays exact. Frame 0 is the static symbol. */
+/** The compact jeo forge mark: a horizontal lobster/crayfish (가재) emblem composed
+ *  of SIMPLE, DISCONNECTED shapes (no connecting strokes) that together read as the
+ *  mascot lying on its side, left→right: the raised pincer CLAWS (집게) as the splayed
+ *  corner wedges (◤◣ left, ◥◢ right), the body carrying the JEO wordmark (J E O), and
+ *  a DNA double-helix woven above and below as a row of crossing nodes (╳ = base-pair
+ *  crossings). gjc-forge aesthetic: clean negative space, geometric symmetry, the
+ *  blue→violet→pink flow gradient applied by renderForgeMark doing the neon glow. The
+ *  lobster identity is carried by the disconnected silhouette; the JEO typography is
+ *  the deliberate lettermark at the core. Width-1 glyphs only (box drawing + geometrics)
+ *  so padding/centering math stays exact. Frame 0 is the static symbol. */
 export const FORGE_MARK_ART: string[] = [
-  "    ◣◣     ◢◢       ",
-  "   ◆══╲   ╱══◆      ",
-  "      ╲◉◉◉╱         ",
-  "       ╲▔▔╱         "
+  "◤    ╳ ╳ ╳ ╳    ◥",
+  "❮     J E O     ▮",
+  "◣    ╳ ╳ ╳ ╳    ◢"
 ];
 
 export const FORGE_MARK_ART_ASCII: string[] = [
-  "    //     \\\\       ",
-  "   o==\\   /==o      ",
-  "      \\ooo/         ",
-  "       \\__/         "
+  "/    x x x x    \\",
+  "<     J E O     |",
+  "\\    x x x x    /"
 ];
 
-/** Claw-snap blink frames for the compact crayfish forge mark: the arms, head and
- *  carapace stay fixed while the two pincer jaws SNAP shut (◣◣/◢◢ open → ◢◣/◢◣
- *  closed), so the crayfish "clicks" its claws. Frame 0 === FORGE_MARK_ART, so a
- *  frameless render is byte-identical to the static symbol. All lines share the
- *  same width and width-1 glyphs. */
+/** Claw-snap blink frames for the compact lobster forge mark: the helix nodes, the
+ *  JEO body and the inner claw/tail glyphs stay fixed while the four splayed pincer
+ *  CORNERS snap (◤◣ / ◥◢ open → ◢◥ / ◣◤ closed), so the lobster "clicks" its claws.
+ *  Frame 0 === FORGE_MARK_ART, so a frameless render is byte-identical to the static
+ *  symbol. All lines share the same width and width-1 glyphs. */
 export const FORGE_MARK_FRAMES: string[][] = [
   FORGE_MARK_ART,
   [
-    "     ◢◣   ◢◣        ",
-    "   ◆══╲   ╱══◆      ",
-    "      ╲◉◉◉╱         ",
-    "       ╲▔▔╱         "
+    "◢    ╳ ╳ ╳ ╳    ◣",
+    "❮     J E O     ▮",
+    "◥    ╳ ╳ ╳ ╳    ◤"
   ]
 ];
 
 export const FORGE_MARK_FRAMES_ASCII: string[][] = [
   FORGE_MARK_ART_ASCII,
   [
-    "     ><   ><        ",
-    "   o==\\   /==o      ",
-    "      \\ooo/         ",
-    "       \\__/         "
+    "\\    x x x x    /",
+    "<     J E O     |",
+    "/    x x x x    \\"
   ]
 ];
 
@@ -466,24 +462,25 @@ export function forgeMarkFrameCount(): number {
   return FORGE_MARK_FRAMES.length;
 }
 
-/** Grand hero variant for the welcome forge box (gjc-style spacious banner): the
- *  same mascot crayfish rendered large and wordless — the two raised pincer claws
- *  (◣◣ / ◢◢, the 집게 feature) on wide arms (◆══╲ ╱══◆), the head with its glowing
- *  eye/terminal cluster (◉ ◉ ◉), and the broad rounded carapace/tail (╲▔▔▔▔▔╱).
- *  Purely symbolic — NO embedded lettering or caption. Width-1 glyphs only so
+/** Grand hero variant for the welcome forge box (gjc-style spacious banner): the same
+ *  horizontal lobster emblem rendered large — the splayed pincer claws as corner wedges
+ *  (◤◣ left, ◥◢ right), the JEO wordmark spaced across the body (J   E   O), and the DNA
+ *  double-helix woven above and below as a wider row of crossing nodes (╳). gjc-forge
+ *  aesthetic: generous negative space + geometric symmetry, with renderForgeMark's
+ *  blue→violet→pink flow gradient supplying the neon glow. The JEO typography is the
+ *  deliberate lettermark; the lobster reads from the disconnected silhouette. Width 29
+ *  (matches the welcome compact↔grand threshold) and width-1 glyphs only so
  *  padding/centering math stays exact. */
 export const FORGE_MARK_ART_GRAND: string[] = [
-  "      ◣◣            ◢◢        ",
-  "    ◆══╲            ╱══◆      ",
-  "        ╲  ◉ ◉ ◉  ╱           ",
-  "         ╲▔▔▔▔▔▔▔▔▔╱          "
+  "◤        ╳ ╳ ╳ ╳ ╳ ╳        ◥",
+  "❮         J   E   O         ▮",
+  "◣        ╳ ╳ ╳ ╳ ╳ ╳        ◢"
 ];
 
 export const FORGE_MARK_ART_GRAND_ASCII: string[] = [
-  "      //            \\\\        ",
-  "    o==\\            /==o      ",
-  "        \\  o o o  /           ",
-  "         \\_________/          "
+  "/        x x x x x x        \\",
+  "<         J   E   O         |",
+  "\\        x x x x x x        /"
 ];