@selesai/code 0.1.19 → 0.2.1

package/dist/defaults/models.json

@@ -1,86 +1,190 @@
 {
   "providers": {
     "tokenin": {
-      "api": "openai-responses",
+      "api": "openai-completions",
       "apiKey": "sk-xxx",
       "baseUrl": "https://lite.andlet.me/v1",
+      "authHeader": true,
+      "compat": {
+        "supportsDeveloperRole": false,
+        "supportsUsageInStreaming": true,
+        "maxTokensField": "max_tokens"
+      },
       "models": [
         {
           "id": "kimi-k2.7-code",
-          "name": "Kimi K2.7-code",
+          "name": "Kimi K2.7 Code",
           "reasoning": true,
           "input": ["text", "image"],
-          "contextWindow": 256000,
-          "maxTokens": 65536,
+          "contextWindow": 262128,
           "cost": {
             "input": 0.95,
             "output": 4,
-            "cacheRead": 0.16,
+            "cacheRead": 0.2375,
             "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": false
           }
         },
         {
-          "id": "glm-5.2",
-          "name": "GLM 5.2",
+          "id": "kimi-k2.6",
+          "name": "Kimi K2.6",
           "reasoning": true,
-          "input": ["text"],
-          "contextWindow": 976000,
-          "maxTokens": 65536,
+          "input": ["text", "image"],
+          "contextWindow": 262128,
+          "cost": {
+            "input": 0.69,
+            "output": 3.22,
+            "cacheRead": 0.1725,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": false
+          }
+        },
+        {
+          "id": "kimi-k2.6-fast",
+          "name": "Kimi K2.6 Fast",
+          "reasoning": false,
+          "input": ["text", "image"],
+          "contextWindow": 262128,
           "cost": {
-            "input": 1.4,
-            "output": 4.4,
-            "cacheRead": 0.26,
+            "input": 0.69,
+            "output": 3.22,
+            "cacheRead": 0.1725,
             "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": false
           }
         },
         {
-          "id": "deepseek-v4-flash",
-          "name": "DeepSeek V4 Flash",
+          "id": "qwen3.5-397b",
+          "name": "Qwen3.5 397B",
           "reasoning": true,
           "input": ["text"],
-          "contextWindow": 1000000,
-          "maxTokens": 384000,
+          "contextWindow": 262128,
           "cost": {
-            "input": 0.14,
-            "output": 0.28,
-            "cacheRead": 0.0028,
+            "input": 0.69,
+            "output": 4.14,
+            "cacheRead": 0.1725,
             "cacheWrite": 0
           },
           "compat": {
-            "requiresReasoningContentOnAssistantMessages": true,
-            "thinkingFormat": "deepseek"
+            "supportsReasoningEffort": false
           }
         },
         {
-          "id": "deepseek-v4-pro",
-          "name": "DeepSeek V4 Pro",
-          "reasoning": true,
+          "id": "qwen3.5-397b-fast",
+          "name": "Qwen3.5 397B Fast",
+          "reasoning": false,
           "input": ["text"],
-          "contextWindow": 1000000,
-          "maxTokens": 384000,
+          "contextWindow": 262128,
           "cost": {
-            "input": 1.74,
-            "output": 3.48,
-            "cacheRead": 0.0145,
+            "input": 0.69,
+            "output": 4.14,
+            "cacheRead": 0.1725,
             "cacheWrite": 0
           },
           "compat": {
-            "requiresReasoningContentOnAssistantMessages": true,
-            "thinkingFormat": "deepseek"
+            "supportsReasoningEffort": false
           }
         },
         {
-          "id": "minimax-m3",
-          "name": "MiniMax M3",
+          "id": "qwen3.6-35b",
+          "name": "Qwen3.6 35B",
           "reasoning": true,
           "input": ["text", "image"],
-          "contextWindow": 512000,
-          "maxTokens": 131072,
+          "contextWindow": 131056,
           "cost": {
-            "input": 0.6,
-            "output": 2.4,
-            "cacheRead": 0.12,
+            "input": 0.29,
+            "output": 1.15,
+            "cacheRead": 0.0725,
             "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": false
+          }
+        },
+        {
+          "id": "qwen3.6-35b-fast",
+          "name": "Qwen3.6 35B Fast",
+          "reasoning": false,
+          "input": ["text", "image"],
+          "contextWindow": 131056,
+          "cost": {
+            "input": 0.29,
+            "output": 1.15,
+            "cacheRead": 0.0725,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": false
+          }
+        },
+        {
+          "id": "glm-5.2",
+          "name": "GLM-5.2",
+          "reasoning": true,
+          "input": ["text"],
+          "contextWindow": 1048560,
+          "cost": {
+            "input": 1.45,
+            "output": 4.5,
+            "cacheRead": 0.3625,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": true
+          }
+        },
+        {
+          "id": "glm-5.2-fast",
+          "name": "GLM-5.2 Fast",
+          "reasoning": false,
+          "input": ["text"],
+          "contextWindow": 1048560,
+          "cost": {
+            "input": 1.45,
+            "output": 4.5,
+            "cacheRead": 0.3625,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": true
+          }
+        },
+        {
+          "id": "glm-5.2-short",
+          "name": "GLM-5.2 Short",
+          "reasoning": true,
+          "input": ["text"],
+          "contextWindow": 199984,
+          "cost": {
+            "input": 1.45,
+            "output": 4.5,
+            "cacheRead": 0.3625,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": true
+          }
+        },
+        {
+          "id": "glm-5.2-short-fast",
+          "name": "GLM-5.2 Short Fast",
+          "reasoning": false,
+          "input": ["text"],
+          "contextWindow": 199984,
+          "cost": {
+            "input": 1.45,
+            "output": 4.5,
+            "cacheRead": 0.3625,
+            "cacheWrite": 0
+          },
+          "compat": {
+            "supportsReasoningEffort": true
           }
         }
       ]

package/dist/defaults/settings.json

@@ -20,27 +20,8 @@
   "packages": [
     "git:github.com/MasuRii/pi-tool-display",
     "git:github.com/omaclaren/pi-markdown-preview",
-    "git:github.com/aliou/pi-guardrails",
-    "git:github.com/nicobailon/pi-intercom",
-      "git:github.com/nicobailon/pi-rewind-hook"
+    "git:github.com/aliou/pi-guardrails"
   ],
-  "powerline": {
-    "preset": "custom",
-    "fixedEditor": true,
-    "fixedEditorPromptGlyph": "",
-    "mouseScroll": true,
-    "customItems": [
-      {
-        "id": "tps",
-        "statusKey": "tps",
-        "position": "left",
-        "prefix": "",
-        "color": "success",
-        "hideWhenMissing": true,
-        "excludeFromExtensionStatuses": true
-      }
-    ]
-  },
   "quietStartup": true,
   "retry": {
     "enabled": true,
@@ -53,13 +34,5 @@
   },
   "theme": "dark",
   "transport": "websocket",
-  "treeFilterMode": "no-tools",
-  "webSearch": {
-    "apiKeys": {
-      "BRAVE_API_KEY": "",
-      "EXA_API_KEY": "",
-      "JINA_API_KEY": "",
-      "TAVILY_API_KEY": ""
-    }
-  }
+  "treeFilterMode": "no-tools"
 }

package/dist/extensions/package.json

@@ -7,14 +7,15 @@
 		"extensions": [
 			"./copy-turn.ts",
 			"./gitignore-guard.ts",
-			"./question.ts",
+			"./tool-error-autofix.ts",
+			"./question",
 			"./rtk.ts",
-			"./tps-tracker.ts",
 			"./tokenin-onboarding.ts",
 			"./undo.ts",
-			"./prototype.ts",
-			"./quick.ts",
+			"./workflow",
 			"./pi-subagents",
+			"./pi-web-agent",
+			"./web-agent-onboarding.ts",
 			"./pi-powerline-footer",
 			"./pi-extension"
 		]

package/dist/extensions/pi-intercom/index.ts

@@ -17,11 +17,11 @@ const SUBAGENT_RESULT_INTERCOM_DELIVERY_EVENT = "subagent:result-intercom-delive
 const INBOUND_FLUSH_DELAY_MS = 200;
 const INBOUND_IDLE_RETRY_MS = 500;
 const DEFAULT_UNNAMED_SESSION_ALIAS_PREFIX = "subagent-chat";
-const SUBAGENT_ORCHESTRATOR_TARGET_ENV = "PI_SUBAGENT_ORCHESTRATOR_TARGET";
-const SUBAGENT_RUN_ID_ENV = "PI_SUBAGENT_RUN_ID";
-const SUBAGENT_CHILD_AGENT_ENV = "PI_SUBAGENT_CHILD_AGENT";
-const SUBAGENT_CHILD_INDEX_ENV = "PI_SUBAGENT_CHILD_INDEX";
-const SUBAGENT_INTERCOM_SESSION_NAME_ENV = "PI_SUBAGENT_INTERCOM_SESSION_NAME";
+const SUBAGENT_ORCHESTRATOR_TARGET_ENV = "SELESAI_SUBAGENT_ORCHESTRATOR_TARGET";
+const SUBAGENT_RUN_ID_ENV = "SELESAI_SUBAGENT_RUN_ID";
+const SUBAGENT_CHILD_AGENT_ENV = "SELESAI_SUBAGENT_CHILD_AGENT";
+const SUBAGENT_CHILD_INDEX_ENV = "SELESAI_SUBAGENT_CHILD_INDEX";
+const SUBAGENT_INTERCOM_SESSION_NAME_ENV = "SELESAI_SUBAGENT_INTERCOM_SESSION_NAME";
 
 interface ChildOrchestratorMetadata {
   orchestratorTarget: string;
@@ -1006,6 +1006,13 @@ export default function piIntercomExtension(pi: ExtensionAPI) {
     syncPresenceIdentity(ctx.sessionManager.getSessionId());
     replyTracker.beginTurn();
   });
+  pi.on("session_info_changed", (_event, ctx) => {
+    if (!getLiveContext(ctx)) {
+      return;
+    }
+    currentSessionId = ctx.sessionManager.getSessionId();
+    syncPresenceIdentity(ctx.sessionManager.getSessionId());
+  });
   pi.on("model_select", (event, ctx) => {
     if (!getLiveContext(ctx)) {
       return;

package/dist/extensions/pi-powerline-footer/index.ts

@@ -43,7 +43,7 @@ import {
   shortcutConflictKey,
   shortcutUsesSuper,
 } from "./shortcuts.ts";
-import { 
+import {
   initVibeManager, 
   onVibeBeforeAgentStart, 
   onVibeAgentStart, 
@@ -59,6 +59,7 @@ import {
   getVibeFileCount,
   generateVibesBatch,
 } from "./working-vibes.ts";
+import { setupTpsTracker } from "./tps.ts";
 
 // ═══════════════════════════════════════════════════════════════════════════
 // Configuration
@@ -1175,6 +1176,8 @@ export default function powerlineFooter(pi: ExtensionAPI) {
   }
 
   // Track session start
+  setupTpsTracker(pi);
+
   pi.on("session_start", async (event, ctx) => {
     shellSession?.dispose();
     shellSession = null;

package/dist/extensions/pi-powerline-footer/tps.ts

@@ -0,0 +1,286 @@
+/**
+ * TPS Tracker — tokens-per-second tracking merged into the powerline footer.
+ *
+ * Writes live and final TPS stats to `ctx.ui.setStatus("tps", ...)`, which the
+ * powerline `extension_statuses` segment surfaces in the status bar (and the
+ * `setExtensionStatus` repaint hook turns into a re-render automatically).
+ *
+ * Tracks:
+ *  - Main model output tokens vs. streaming wall time (excludes first-token
+ *    latency and tool execution gaps).
+ *  - pi-subagents child output tokens via tool_execution_* events, aggregated
+ *    into a combined "main + subagents" TPS at agent_end.
+ */
+
+import type { ExtensionAPI } from "@earendil-works/pi-coding-agent";
+
+export function setupTpsTracker(pi: ExtensionAPI): void {
+	/** Timestamp when the current assistant message event started. Used as a fallback. */
+	let messageStart: number | null = null;
+	/** Timestamp of the first streamed output delta for the current assistant message. */
+	let streamStart: number | null = null;
+	/** Estimated streamed output tokens for live display before providers report final usage. */
+	let estimatedStreamedTokens = 0;
+	/** Cumulative official output tokens across all assistant messages in this agent run. */
+	let totalOutputTokens = 0;
+	/** Cumulative time (ms) spent actually streaming output deltas (excludes tool execution and first-token latency). */
+	let totalStreamMs = 0;
+	/** Subagent tool-call start times, used to include pi-subagents child output in aggregate TPS. */
+	const subagentStarts = new Map<string, number>();
+	/** Latest cumulative foreground subagent tokens seen during tool_execution_update. */
+	const subagentLiveTokens = new Map<string, number>();
+	/** Cumulative subagent output tokens in this agent run. */
+	let totalSubagentTokens = 0;
+	/** Cumulative subagent wall time (ms) in this agent run. */
+	let totalSubagentMs = 0;
+	/** Completed subagent call summaries in this agent run. */
+	const subagentSummaries: Array<{ tokens: number; ms: number; tps: number }> = [];
+
+	function asRecord(value: unknown): Record<string, unknown> | null {
+		return value && typeof value === "object" && !Array.isArray(value) ? value as Record<string, unknown> : null;
+	}
+
+	function numberValue(value: unknown): number {
+		return typeof value === "number" && Number.isFinite(value) ? value : 0;
+	}
+
+	function getDetailsResult(value: unknown): Record<string, unknown> {
+		const result = asRecord(value);
+		const details = asRecord(result?.details);
+		return details ?? result ?? {};
+	}
+
+	function getResults(value: unknown): Record<string, unknown>[] {
+		const details = getDetailsResult(value);
+		return Array.isArray(details.results)
+			? details.results.map(asRecord).filter((item): item is Record<string, unknown> => !!item)
+			: [];
+	}
+
+	function generatedTokensFromUsage(usage: Record<string, unknown> | null): number {
+		return numberValue(usage?.output ?? usage?.outputTokens ?? usage?.completionTokens ?? usage?.completion_tokens);
+	}
+
+	function generatedTokensFromMessage(message: { usage?: unknown; content?: unknown }): number {
+		const usageTokens = generatedTokensFromUsage(asRecord(message.usage));
+		if (usageTokens > 0) return usageTokens;
+
+		const content = Array.isArray(message.content) ? message.content : [];
+		const chars = content.reduce((sum, block) => {
+			const item = asRecord(block);
+			if (!item) return sum;
+			if (item.type === "text") return sum + String(item.text ?? "").length;
+			if (item.type === "thinking") return sum + String(item.thinking ?? "").length;
+			if (item.type === "toolCall") return sum + String(item.name ?? "").length + JSON.stringify(item.arguments ?? {}).length;
+			return sum;
+		}, 0);
+		return Math.ceil(chars / 4);
+	}
+
+	function outputTokensFromResult(value: unknown): number {
+		const results = getResults(value);
+		return results.reduce((sum, result) => sum + generatedTokensFromUsage(asRecord(result.usage)), 0);
+	}
+
+	function liveTokensFromResult(value: unknown): number {
+		return subagentProgressItems(value).reduce((sum, item) => sum + item.tokens, 0);
+	}
+
+	function subagentProgressItems(value: unknown): Array<{ name: string; tokens: number; status?: string }> {
+		const details = getDetailsResult(value);
+		const progress = Array.isArray(details.progress) ? details.progress : undefined;
+		const resultItems = getResults(value).map((result, index) => {
+			const progress = asRecord(result.progress);
+			const usage = asRecord(result.usage);
+			return {
+				name: String(result.agent || progress?.agent || `agent${index + 1}`),
+				tokens: generatedTokensFromUsage(usage) || generatedTokensFromUsage(progress),
+				status: typeof progress?.status === "string" ? progress.status : undefined,
+			};
+		});
+		if (!progress || resultItems.some((item) => item.tokens > 0)) return resultItems;
+
+		return progress.map((entry, index) => {
+			const item = asRecord(entry) ?? {};
+			return {
+				name: String(item.agent || `agent${index + 1}`),
+				tokens: generatedTokensFromUsage(item),
+				status: typeof item.status === "string" ? item.status : undefined,
+			};
+		});
+	}
+
+	function subagentModeLabel(value: unknown): string {
+		const details = getDetailsResult(value);
+		const mode = typeof details.mode === "string" ? details.mode : "subagents";
+		return mode === "parallel" ? "parallel" : mode === "chain" ? "chain" : "subagent";
+	}
+
+	pi.on("agent_start", async (_event, ctx) => {
+		totalOutputTokens = 0;
+		totalStreamMs = 0;
+		totalSubagentTokens = 0;
+		totalSubagentMs = 0;
+		subagentSummaries.length = 0;
+		messageStart = null;
+		streamStart = null;
+		estimatedStreamedTokens = 0;
+		subagentStarts.clear();
+		subagentLiveTokens.clear();
+		const theme = ctx.ui.theme;
+		ctx.ui.setStatus("tps", theme.fg("success", "generating..."));
+	});
+
+	pi.on("message_start", async (event) => {
+		if (event.message.role !== "assistant") return;
+		messageStart = Date.now();
+		streamStart = null;
+		estimatedStreamedTokens = 0;
+	});
+
+	pi.on("message_update", async (event, ctx) => {
+		if (event.message.role !== "assistant") return;
+
+		const streamEvent = event.assistantMessageEvent;
+		const isOutputDelta =
+			streamEvent.type === "text_delta" ||
+			streamEvent.type === "thinking_delta" ||
+			streamEvent.type === "toolcall_delta";
+
+		if (!isOutputDelta) return;
+
+		const now = Date.now();
+		streamStart ??= now;
+		estimatedStreamedTokens += Math.max(0, streamEvent.delta.length / 4);
+
+		const elapsed = (now - streamStart) / 1000;
+		const officialTokens = generatedTokensFromUsage(asRecord(event.message.usage));
+		const currentTokens = officialTokens > 0 ? officialTokens : estimatedStreamedTokens;
+
+		if (elapsed > 0 && currentTokens > 0) {
+			const tps = Math.round(currentTokens / elapsed);
+			const tokenLabel = officialTokens > 0
+				? `${officialTokens} tok`
+				: `~${Math.round(estimatedStreamedTokens)} tok`;
+			const theme = ctx.ui.theme;
+			ctx.ui.setStatus(
+				"tps",
+				`${theme.fg("accent", `${tps} tok/s`)}`,
+			);
+		}
+	});
+
+	pi.on("message_end", async (event) => {
+		if (event.message.role !== "assistant") return;
+
+		const messageTokens = generatedTokensFromMessage(event.message);
+		const timingStart = streamStart ?? messageStart;
+		if (!timingStart || messageTokens <= 0) {
+			messageStart = null;
+			streamStart = null;
+			estimatedStreamedTokens = 0;
+			return;
+		}
+
+		totalOutputTokens += messageTokens;
+		totalStreamMs += Math.max(0, Date.now() - timingStart);
+
+		messageStart = null;
+		streamStart = null;
+		estimatedStreamedTokens = 0;
+	});
+
+	pi.on("tool_execution_start", async (event) => {
+		if (event.toolName !== "subagent") return;
+		subagentStarts.set(event.toolCallId, Date.now());
+	});
+
+	pi.on("tool_execution_update", async (event, ctx) => {
+		if (event.toolName !== "subagent") return;
+
+		const start = subagentStarts.get(event.toolCallId);
+		if (!start) return;
+
+		const items = subagentProgressItems(event.partialResult);
+		const tokens = items.reduce((sum, item) => sum + item.tokens, 0);
+		subagentLiveTokens.set(event.toolCallId, tokens);
+		const elapsed = (Date.now() - start) / 1000;
+		if (tokens <= 0 || elapsed <= 0) return;
+
+		const tps = Math.round(tokens / elapsed);
+		const mode = subagentModeLabel(event.partialResult);
+		const agentSummary = items
+			.filter((item) => item.tokens > 0 || item.status === "running")
+			.map((item) => `${item.name} ${Math.round(item.tokens / elapsed)} t/s`)
+			.join(" ");
+		const theme = ctx.ui.theme;
+		ctx.ui.setStatus("tps", `${theme.fg("accent", `main+${mode} ${tps} t/s`)} ${theme.fg("success", `(${agentSummary || `${tps} t/s`})`)}`);
+	});
+
+	pi.on("tool_execution_end", async (event, ctx) => {
+		if (event.toolName !== "subagent") return;
+
+		const start = subagentStarts.get(event.toolCallId);
+		subagentStarts.delete(event.toolCallId);
+		const liveTokens = subagentLiveTokens.get(event.toolCallId) ?? liveTokensFromResult(event.result);
+		subagentLiveTokens.delete(event.toolCallId);
+		if (!start || event.isError) return;
+
+		const outputTokens = outputTokensFromResult(event.result);
+		const displayTokens = outputTokens || liveTokens;
+		const mode = subagentModeLabel(event.result);
+		const elapsedMs = Math.max(0, Date.now() - start);
+		if (displayTokens <= 0 || elapsedMs <= 0) return;
+		const elapsedSeconds = elapsedMs / 1000;
+		const items = subagentProgressItems(event.result);
+		const agentSummary = items
+			.filter((item) => item.tokens > 0)
+			.map((item) => `${item.name} ${Math.round(item.tokens / elapsedSeconds)} t/s`)
+			.join(" ");
+
+		totalOutputTokens += displayTokens;
+		totalStreamMs += elapsedMs;
+		totalSubagentTokens += displayTokens;
+		totalSubagentMs += elapsedMs;
+
+		const tps = Math.round(displayTokens / elapsedSeconds);
+		subagentSummaries.push({ tokens: displayTokens, ms: elapsedMs, tps });
+		const theme = ctx.ui.theme;
+		ctx.ui.notify(`${theme.fg("success", "✓")} ${theme.fg("accent", `${tps} t/s`)}`, "info");
+		ctx.ui.setStatus(
+			"tps",
+			`${theme.fg("accent", `main+${mode} ${tps} t/s`)} ${theme.fg("success", `(${agentSummary || `${tps} t/s`})`)}`,
+		);
+	});
+
+	pi.on("agent_end", async (_event, ctx) => {
+		const elapsed = totalStreamMs / 1000;
+		const tps = totalOutputTokens > 0 && elapsed > 0 ? Math.round(totalOutputTokens / elapsed) : 0;
+
+		const theme = ctx.ui.theme;
+		const icon = theme.fg("success", "✓");
+		const tpsLabel = tps > 0
+			? theme.fg("accent", `${tps} t/s`)
+			: theme.fg("success", "N/A");
+		const mainTokens = totalOutputTokens - totalSubagentTokens;
+		const mainMs = totalStreamMs - totalSubagentMs;
+		const mainTps = mainTokens > 0 && mainMs > 0 ? Math.round(mainTokens / (mainMs / 1000)) : 0;
+		const subagentTps = totalSubagentTokens > 0 && totalSubagentMs > 0
+			? Math.round(totalSubagentTokens / (totalSubagentMs / 1000))
+			: 0;
+		const mainDetail = mainTps > 0 ? theme.fg("success", ` | main ${mainTps} t/s`) : "";
+		const subagentDetail = subagentTps > 0
+			? theme.fg("success", ` | subagents ${subagentTps} t/s (${subagentSummaries.length} call${subagentSummaries.length === 1 ? "" : "s"})`)
+			: "";
+
+		ctx.ui.notify(`${icon} total ${tpsLabel}${mainDetail}${subagentDetail}`, "info");
+		const finalMainStatus = mainTps > 0 ? `main ${mainTps} t/s` : "main N/A";
+		const finalSubagentStatus = totalSubagentTokens > 0 && totalSubagentMs > 0
+			? `subagents ${Math.round(totalSubagentTokens / (totalSubagentMs / 1000))} t/s`
+			: "";
+		ctx.ui.setStatus(
+			"tps",
+			`${theme.fg("accent", `done ${tps} t/s`)} ${theme.fg("success", `(${[finalMainStatus, finalSubagentStatus].filter(Boolean).join(" ")})`)}`,
+		);
+	});
+}