@oh-my-pi/pi-coding-agent 15.10.9 → 15.10.11

package/src/prompts/tools/irc.md

@@ -4,30 +4,30 @@ Sends short text messages to other live agents in this process and receives thei
 - The main agent is addressable as `Main`. Subagents reuse their task id (e.g. `AuthLoader`, or `AuthLoader-2` when the name repeats).
 - `op: "list"` returns the current set of visible peers. Use it before sending if you are not sure who is live.
 - `op: "send"` delivers `message` to `to`. `to` may be a specific id or `"all"` to broadcast.
-- The recipient generates the reply via an ephemeral side-channel turn that uses their current model, system prompt, and history — it does **not** wait for the recipient's main loop to be free, so it is safe to IRC an agent that is currently inside a long-running tool call.
-- The exchange (incoming question + auto-reply) is queued for injection into the recipient's persisted history; the recipient sees it on its next turn and can follow up if needed.
+- Replies are generated on a side channel that does not wait for the recipient's main loop, so it is safe to IRC an agent that is mid tool call.
+- The exchange (question + auto-reply) is injected into the recipient's history; they see it on their next turn and can follow up.
 </instruction>
 
 <when_to_use>
 You SHOULD reach for `irc` proactively when continuing alone is wasteful or wrong. When in doubt, prefer messaging.
-- **Unexpected state.** You hit something the original task did not describe — a missing file, a config that contradicts the assignment, an API behaving differently than you were told, a tool failing in a way that suggests the spec is wrong. DM `Main` (or the spawning agent) for guidance instead of guessing.
-- **Blocked by another agent.** A peer holds the file/branch/resource you need, has already started the change you are about to make, or owns a decision you depend on. DM that peer (or broadcast to discover who) before duplicating or stepping on work.
-- **Decision points outside your scope.** A genuine fork in the road that the assignment did not pre-decide (e.g. which of two viable APIs to use, whether to refactor adjacent code). Ask the requester rather than picking unilaterally.
-- **Coordination opportunities.** You realize a peer's in-flight work would benefit from yours, or vice-versa.
+- **Unexpected state.** The task did not describe what you found — missing file, config contradicting the assignment, API or tool behaving differently than told. DM `Main` (or the spawning agent) instead of guessing.
+- **Blocked by another agent.** A peer holds the file/branch/resource you need, started the change you are about to make, or owns a decision you depend on. DM that peer (or broadcast to discover who) before duplicating work.
+- **Decision points outside your scope.** A genuine fork the assignment did not pre-decide (e.g. which of two viable APIs, whether to refactor adjacent code). Ask the requester rather than picking unilaterally.
+- **Coordination opportunities.** A peer's in-flight work would benefit from yours, or vice-versa.
 
-Do **not** use `irc` for: routine progress updates, things you can verify with a tool call, or questions whose answer is already in your assignment / repo / docs.
+NEVER use `irc` for: routine progress updates, things a tool call can verify, or questions already answered by your assignment / repo / docs.
 </when_to_use>
 
 <etiquette>
 These rules apply to both sending and replying.
-- **Plain prose only.** Do not send structured JSON status payloads (e.g. `{"type":"task_completed",…}`). Write a normal sentence: "Done with the auth refactor — left a TODO in `src/server/auth.ts` for the rate limiter."
-- **Do not quote the message you are replying to.** The sender already saw it; the TUI already renders it. Lead with the answer.
-- **Use IRC, not terminal tools, to learn about peers.** Do not `grep` artifacts, read other sessions' JSONL files, or shell-poke around to figure out what another agent is doing. DM them — they have the live answer and you do not.
-- **One round-trip is enough.** Replies arrive synchronously when the recipient is reachable. Do not follow up with "did you get my message?" — they did. If `delivered` is empty or the result was `failed`, the peer is unavailable; move on or report the blocker, do not retry in a loop.
-- **Stay terse.** A DM is a chat message, not a memo. One question per send when you can. Share file paths and artifacts via `local://` / `memory://` / `artifact://` URLs instead of pasting blobs.
-- **Address peers by id.** Use the exact id from `op: "list"` (e.g. `AuthLoader`, `Main`). Do not invent friendly names.
-- **Do not IRC for things a tool would answer.** If a `read`, `grep`, or build command would resolve the question, do that first.
-- **When you receive an IRC message, answer it before continuing.** The recipient injects the question + your auto-reply into your history; address it directly, do not repeat it back to the user.
+- **Plain prose only.** NEVER send structured JSON status payloads (e.g. `{"type":"task_completed",…}`). Write a normal sentence: "Done with the auth refactor — left a TODO in `src/server/auth.ts` for the rate limiter."
+- **NEVER quote the message you are replying to.** Lead with the answer.
+- **Use IRC, not terminal tools, to learn about peers.** NEVER `grep` artifacts, read other sessions' JSONL files, or shell-poke to figure out what another agent is doing. DM them.
+- **One round-trip is enough.** Replies arrive synchronously when the recipient is reachable. NEVER follow up with "did you get my message?". If `delivered` is empty or the result was `failed`, the peer is unavailable — move on or report the blocker; NEVER retry in a loop.
+- **Stay terse.** A DM is a chat message, not a memo. One question per send. Share file paths and artifacts via `local://` / `memory://` / `artifact://` URLs instead of pasting blobs.
+- **Address peers by id.** Use the exact id from `op: "list"` (e.g. `AuthLoader`, `Main`). NEVER invent friendly names.
+- **NEVER IRC for things a tool would answer.** If a `read`, `grep`, or build command resolves the question, do that first.
+- **Answer incoming IRC messages before continuing.** Address the question directly; do not repeat it back to the user.
 </etiquette>
 
 <output>

package/src/prompts/tools/lsp.md

@@ -37,6 +37,6 @@ Interacts with Language Server Protocol servers for code intelligence.
 
 <critical>
 - You MUST use `lsp` for symbol-aware operations (rename, find references, go to definition/implementation, code actions) whenever a language server is available — it is safer and more accurate than text-based alternatives.
-- You NEVER perform cross-file renames with `ast_edit`, `sed`, `rsed`, or manual edits when `lsp` `rename` can do it. Text-based renames miss shadowing, re-exports, and usages in other files.
-- Prefer `lsp` `code_actions` for imports, quick-fixes, and refactors the language server already knows how to apply.
+- You NEVER perform cross-file renames with `ast_edit`, `sed`, or manual edits when `lsp` `rename` can do it. Text-based renames miss shadowing, re-exports, and usages in other files.
+- You SHOULD use `lsp` `code_actions` for imports, quick-fixes, and refactors the language server already knows how to apply.
 </critical>

package/src/prompts/tools/patch.md

@@ -5,7 +5,7 @@ Patches files given diff hunks. Primary tool for existing-file edits.
 - `@@` — bare header when context lines unique
 - `@@ $ANCHOR` — anchor copied verbatim from file (full line or unique substring)
 **Anchor Selection:**
-1. Otherwise choose highly specific anchor copied from file:
+1. Prefer bare `@@` when context lines alone are unique; otherwise choose highly specific anchor copied from file:
    - full function signature
    - class declaration
    - unique string literal/error message
@@ -47,7 +47,7 @@ Returns success/failure; on failure, error message indicates:
 - You NEVER use anchors as comments (no line numbers, location labels, placeholders like `@@ @@`)
 - You NEVER place new lines outside the intended block
 - If edit fails or breaks structure, you MUST re-read the file and produce a new patch from current content — you NEVER retry the same diff
-- NEVER use edit to fix indentation, whitespace, or reformat code. Formatting is a single command run once at the end (`bun fmt`, `cargo fmt`, `prettier —write`, etc.)—not N individual edits. If you see inconsistent indentation after an edit, leave it; the formatter will fix all of it in one pass.
+- NEVER use edit to fix indentation, whitespace, or reformat code. Formatting is a single command run once at the end (`bun fmt`, `cargo fmt`, `prettier --write`, etc.) — not N individual edits. If you see inconsistent indentation after an edit, leave it; the formatter will fix all of it in one pass.
 </critical>
 
 <examples>

package/src/prompts/tools/read.md

@@ -8,7 +8,7 @@ Read files, directories, archives, SQLite databases, images, documents, internal
 
 ## Parameters
 
-- `path` — required. Local path, internal URI (`skill://`, `agent://`, `artifact://`, `memory://`, `rule://`, `local://`, `vault://`, `mcp://`), or URL. Append `:<sel>` for line ranges, raw mode, or special modes (e.g. `src/foo.ts:50-200`, `src/foo.ts:raw`, `db.sqlite:users:42`).
+- `path` — required. Local path, internal URI (`skill://`, `agent://`, `artifact://`, `memory://`, `rule://`, `local://`, `vault://`, `mcp://`, `omp://`, `issue://`, `pr://`), or URL. Append `:<sel>` for line ranges, raw mode, or special modes (e.g. `src/foo.ts:50-200`, `src/foo.ts:raw`, `db.sqlite:users:42`).
 
 ## Selectors
 
@@ -74,12 +74,11 @@ For `.sqlite`, `.sqlite3`, `.db`, `.db3`:
 
 # Internal URIs
 
-`skill://<name>`, `agent://<id>`, `artifact://<id>`, `memory://root`, `rule://<name>`, `local://<name>.md`, `vault://<vault>/<path>`, `mcp://<uri>` resolve transparently and accept the same line selectors as filesystem paths. Use `artifact://<id>` to recover full output that a previous bash/eval/tool result spilled or truncated.
+`skill://<name>`, `agent://<id>`, `artifact://<id>`, `memory://root`, `rule://<name>`, `local://<name>.md`, `vault://<vault>/<path>`, `mcp://<uri>`, `omp://<doc>.md`, `issue://<N>`, and `pr://<N>` resolve transparently and accept the same line selectors as filesystem paths. Use `artifact://<id>` to recover full output that a previous bash/eval/tool result spilled or truncated.
 
 <critical>
 - You MUST use `read` for every file, directory, archive, and URL inspection. `cat`, `head`, `tail`, `less`, `more`, `ls`, `tar`, `unzip`, `curl`, `wget` are FORBIDDEN — any such bash call is a bug, regardless of how short or convenient it looks.
 - You MUST prefer `read` over a browser/puppeteer tool for URL content; only reach for a browser when `read` cannot deliver reasonable content.
 - For line ranges, append the selector to `path` (`path="src/foo.ts:50-200"`, `path="src/foo.ts:50+150"`). NEVER substitute `sed -n`, `awk NR`, or `head`/`tail` pipelines.
-- Summary footer says `read <path>:raw …`? Re-issue the exact selector it names. NEVER guess what's inside `..` / `…` markers — they carry no content.
-- You MAY combine selectors with URL reads and internal URIs; both paginate the cached resolved output.
+- Summary footer names ranges to re-read? Re-issue ONLY the ranges you need via the multi-range selector. NEVER guess what's inside `..` / `…` markers — they carry no content.
 </critical>

package/src/prompts/tools/recall.md

@@ -2,4 +2,4 @@ Search long-term memory for relevant information. Returns raw matching entries r
 
 Use proactively — before answering questions about past conversations, user preferences, project decisions, or any topic where prior context would help accuracy. When in doubt, recall first.
 
-Prefer `recall` when you need specific facts or entries. Use `reflect` instead when you need a synthesised answer across many memories.
+Prefer `recall` when you need specific facts or entries. Use `reflect` instead when you need a synthesized answer across many memories.

package/src/prompts/tools/reflect.md

@@ -1,4 +1,4 @@
-Generate a synthesised answer by reasoning over long-term memory. Unlike `recall`, `reflect` blends relevant memories into a coherent response.
+Generate a synthesized answer by reasoning over long-term memory. Unlike `recall`, `reflect` blends relevant memories into a coherent response.
 
 Use for open-ended questions spanning many stored facts: "What do you know about this user?", "Summarize project decisions.", "What are my preferences for X?"
 

package/src/prompts/tools/render-mermaid.md

@@ -3,7 +3,7 @@ Convert Mermaid graph source into ASCII diagram output.
 Parameters:
 - `mermaid` (required): Mermaid graph text to render.
 - `config` (optional): JSON render configuration (spacing and layout options).
+
 Behavior:
 - Returns ASCII diagram text.
-- Saves full output to `artifact://<id>` when storage available.
-- Returns error when Mermaid input invalid or rendering fails.
+- Saves full output to `artifact://<id>`.

package/src/prompts/tools/replace.md

@@ -16,21 +16,15 @@ Returns success/failure status. On success, file modified in place with replacem
 </critical>
 
 <bash-alternatives>
-Replace for content-addressed changes—you identify \_what* to change by its text.
+Replace is content-addressed — you identify *what* to change by its text.
 
-For position-addressed or pattern-addressed changes, bash more efficient:
+For pattern-addressed bulk changes, bash is more efficient:
 
 |Operation|Command|
 |---|---|
-|Append to file|`cat >> file <<'EOF'`…`EOF`|
-|Prepend to file|`{ cat - file; } <<'EOF' > tmp && mv tmp file`|
-|Delete lines N-M|`sed -i 'N,Md' file`|
-|Insert after line N|`sed -i 'Na\text' file`|
 |Regex replace|`sd 'pattern' 'replacement' file`|
 |Bulk replace across files|`sd 'pattern' 'replacement' **/*.ts`|
-|Copy lines N-M to another file|`sed -n 'N,Mp' src >> dest`|
-|Move lines N-M to another file|`sed -n 'N,Mp' src >> dest && sed -i 'N,Md' src`|
 
-Use Replace when _content itself_ identifies location.
-Use bash when _position_ or _pattern_ identifies what to change.
+Use Replace when _content itself_ identifies location; use `ast_edit` for structure-aware codemods.
+NEVER use `sed -i`/`perl -i`/heredoc redirection for edits — those calls are blocked; use this tool or `write`.
 </bash-alternatives>

package/src/prompts/tools/rewind.md

@@ -3,9 +3,9 @@ End an active checkpoint. Rewind context to it, replacing intermediate explorati
 Call immediately after `checkpoint`-started investigative work.
 
 Requirements:
-- `report` is REQUIRED and must be concise, factual, and actionable.
+- `report` is REQUIRED and MUST be concise, factual, and actionable.
 - Include key findings, decisions, and any unresolved risks.
-- Do not include raw scratch logs unless essential.
+- AVOID raw scratch logs unless essential.
 - You MUST call this before yielding if a checkpoint is active.
 
 Behavior:

package/src/prompts/tools/search-tool-bm25.md

@@ -15,21 +15,13 @@ Input:
 - `limit` — optional maximum number of tools to return and activate (default `8`)
 
 Behavior:
-- Searches hidden tool metadata using BM25-style relevance ranking
 - Matches against tool name, label, server name, description/summary, and input schema keys
 - Activates the top matching tools for the rest of the current session
 - Repeated searches add to the active tool set; they do not remove earlier selections
 - Newly activated tools become available before the next model call in the same overall turn
 
 Notes:
-Start with `limit` 5–10 if unsure.
-- `query` is matched against tool metadata fields:
-  - `name`
-  - `label`
-  - `server_name` (MCP tools)
-  - `mcp_tool_name` (MCP tools)
-  - `description` / `summary`
-  - input schema property keys (`schema_keys`)
+- Start with `limit` 5–10 if unsure.
 
 Not for repository/file/code search. Tool discovery only.
 

package/src/prompts/tools/search.md

@@ -20,6 +20,5 @@ Searches files using powerful regex matching.
 <critical>
 - You MUST use the built-in `search` tool for any content search. NEVER shell out to `grep`, `rg`, `ripgrep`, `ag`, `ack`, `git grep`, `awk`, `sed`-for-search, or any other CLI search via Bash — even for a single match, even "just to check quickly", even piped through other commands.
 - Bash `grep`/`rg` loses `.gitignore` semantics, bypasses result limits, and wastes tokens. The `search` tool is faster, structured, and already wired into the workspace — there is no scenario where Bash search is preferable.
-- If you catch yourself typing `grep`, `rg`, or `| grep` in a Bash command, stop and re-issue the lookup through the `search` tool instead.
 - If the search is open-ended, requiring multiple rounds, you MUST use the Task tool with the explore subagent instead of chaining `search` calls yourself.
 </critical>

package/src/prompts/tools/ssh.md

@@ -1,9 +1,5 @@
 Runs commands on remote hosts.
 
-<instruction>
-You MUST build commands from the reference below
-</instruction>
-
 <commands>
 **linux/bash, linux/zsh, macos/bash, macos/zsh** — Unix-like:
 - Files: `ls`, `cat`, `head`, `tail`, `grep`, `find`

package/src/prompts/tools/task.md

@@ -31,10 +31,9 @@ Subagents have no conversation history. Every fact, file path, and direction the
 
 <rules>
 - **Maximize batch width.** Spawn the widest parallel set the work decomposes into. NEVER spawn a single-task batch for divisible work, or defer work that could have been concurrent.
-- NEVER assign tasks to run project-wide build/test/lint. Caller verifies after the batch.
-- **Subagents do not verify, lint, or format.** Every assignment MUST instruct the subagent to skip all gates and formatters. You run them once at the end across the union of changed files — avoids redundant runs and racing formatter passes.
+- **Subagents do not verify, lint, or format.** Every assignment MUST instruct the subagent to skip all gates, formatters, and project-wide build/test/lint. You run them once at the end across the union of changed files — avoids redundant runs and racing formatter passes.
 - No globs, no "update all", no package-wide scope. Fan out.
-- Do not concern yourself with how agents might overlap on certain actions. Never use it as an excuse to go slower: they can resolve collisions in real-time with the harness facilities.
+- NEVER slow down or serialize because tasks might overlap on some files. Agents resolve collisions among themselves in real time.
 - Pass large payloads via `local://<path>` URIs, not inline. {{#if contextEnabled}} (other than the context){{/if}}
 {{#if contextEnabled}}- Put shared constraints in `context` once; do not duplicate across assignments.{{/if}}
 - Prefer agents that investigate **and** edit in one pass; only spin a read-only discovery step when affected files are genuinely unknown.

package/src/prompts/tools/todo.md

@@ -2,7 +2,7 @@
 
 Manages a phased task list. Pass `ops`: a flat array of operations.
 The next pending task is auto-promoted to `in_progress` after each completion.
-Allowed `op` values are only `init`, `start`, `done`, `drop`, `rm`, `append`, and `note`. `pending` is a task status, not an `op`; leave not-yet-started tasks implicit in `init`/`append` lists.
+Allowed `op` values are only `init`, `start`, `done`, `drop`, `rm`, `append`, `note`, and `view`. `pending` is a task status, not an `op`; leave not-yet-started tasks implicit in `init`/`append` lists.
 
 ## Operations
 
@@ -12,9 +12,10 @@ Allowed `op` values are only `init`, `start`, `done`, `drop`, `rm`, `append`, an
 |`start`|`task`|Mark in progress|
 |`done`|`task` or `phase`|Mark completed|
 |`drop`|`task` or `phase`|Mark abandoned|
-|`rm`|`task` or `phase`|Remove|
+|`rm`|`task` or `phase` (optional)|Remove task or phase's tasks; omit both to clear the entire list|
 |`append`|`phase`, `items: string[]`|Append tasks to `phase`; lazily creates phase|
 |`note`|`task`, `text`|Append a note to a task. Reminders for future-you only.|
+|`view`|—|Read-only: echo the current list without modifying it|
 
 ## Anatomy
 - **Task content**: 5–10 words, what is being done, not how. Used as the task identifier — unique.
@@ -25,6 +26,7 @@ Allowed `op` values are only `init`, `start`, `done`, `drop`, `rm`, `append`, an
 - Complete phases in order.
 - On blockers, `append` a new task to the active phase to unblock yourself, or `drop`.
 - `task` and `phase` fields reference content/name verbatim; keep them stable once introduced.
+- Lost track of exact task text? `view` echoes the full list — NEVER guess content from memory; a mismatched `task` string is an error.
 
 ## When to create a list
 - Task requires 3+ distinct steps
@@ -35,6 +37,8 @@ Allowed `op` values are only `init`, `start`, `done`, `drop`, `rm`, `append`, an
 <examples>
 # Initial setup (multi-phase)
 `{"ops":[{"op":"init","list":[{"phase":"Foundation","items":["Scaffold crate","Wire workspace"]},{"phase":"Auth","items":["Port credential store","Wire OAuth providers"]},{"phase":"Verification","items":["Run cargo test"]}]}]}`
+# View current state (read-only)
+`{"ops":[{"op":"view"}]}`
 # Initial setup (single phase)
 `{"ops":[{"op":"init","list":[{"phase":"Implementation","items":["Apply fix","Run tests"]}]}]}`
 # Complete one task

package/src/sdk.ts

@@ -19,6 +19,7 @@ import {
 	getOpenAICodexTransportDetails,
 	prewarmOpenAICodexResponses,
 } from "@oh-my-pi/pi-ai/providers/openai-codex-responses";
+import { DEFAULT_MODEL_PER_PROVIDER } from "@oh-my-pi/pi-catalog/provider-models";
 import type { Component } from "@oh-my-pi/pi-tui";
 import {
 	$env,
@@ -41,7 +42,6 @@ import { createApiKeyResolver } from "./config/api-key-resolver";
 import { shouldEnableAppendOnlyContext } from "./config/append-only-context-mode";
 import { ModelRegistry } from "./config/model-registry";
 import {
-	defaultModelPerProvider,
 	formatModelString,
 	getModelMatchPreferences,
 	parseModelPattern,
@@ -531,11 +531,18 @@ function resolveSnapshotTtlMs(): number {
  * override to re-mint access tokens when needed.
  */
 export async function discoverAuthStorage(agentDir: string = getDefaultAgentDir()): Promise<AuthStorage> {
-	const brokerConfig = await resolveAuthBrokerConfig();
+	const brokerConfigPromise = resolveAuthBrokerConfig();
+	const cachePath = getAuthBrokerSnapshotCachePath();
+	// Warm the encrypted snapshot cache into the page cache while the broker
+	// config resolves (it may shell out for a `!command` token). Decryption
+	// needs the resolved token, so the real cache read cannot start earlier.
+	void Bun.file(cachePath)
+		.arrayBuffer()
+		.catch(() => undefined);
+	const brokerConfig = await brokerConfigPromise;
 	if (brokerConfig) {
 		const client = new AuthBrokerClient({ url: brokerConfig.url, token: brokerConfig.token });
 		const ttlMs = resolveSnapshotTtlMs();
-		const cachePath = getAuthBrokerSnapshotCachePath();
 		const persist =
 			ttlMs > 0
 				? (snapshot: SnapshotResponse): void => {
@@ -1730,7 +1737,7 @@ export async function createAgentSession(options: CreateAgentSessionOptions = {}
 			// the winning provider (e.g. anthropic's claude-3-5-sonnet-20240620)
 			// instead of the intended provider default (claude-sonnet-4-6). Mirrors
 			// findInitialModel's precedence.
-			for (const [provider, defaultId] of Object.entries(defaultModelPerProvider)) {
+			for (const [provider, defaultId] of Object.entries(DEFAULT_MODEL_PER_PROVIDER)) {
 				const preferred = fallbackCandidates.find(
 					candidate => candidate.provider === provider && candidate.id === defaultId,
 				);
@@ -2342,7 +2349,8 @@ export async function createAgentSession(options: CreateAgentSessionOptions = {}
 		}
 
 		if (model?.api === "openai-codex-responses") {
-			const codexModel = model;
+			// `.api` equality doesn't narrow the generic; the guard makes this cast sound.
+			const codexModel = model as Model<"openai-codex-responses">;
 			const codexTransport = getOpenAICodexTransportDetails(codexModel, {
 				sessionId: providerSessionId,
 				baseUrl: codexModel.baseUrl,
@@ -2373,12 +2381,17 @@ export async function createAgentSession(options: CreateAgentSessionOptions = {}
 		}
 
 		// Start LSP warmup in the background so startup does not block on language server initialization.
-		// Print/script invocations (`hasUI=false`) don't render the warmup status indicator AND typically
-		// finish before LSP servers would have stabilized — warming them just spends CPU parsing big
-		// `initialize` responses concurrently with the LLM stream consumer, jittering perceived latency.
-		// Tools that need an LSP server still spin one up on demand through `getOrCreateClient`.
+		// With `lsp.lazy` (the default) the warmup is skipped: recognized servers are still discovered and
+		// surfaced in the UI as "available", but cold-start on first use — the lsp tool or an edit/write
+		// touching a matching file type — through `getOrCreateClient`.
+		// Print/script invocations (`hasUI=false`) skip it regardless: they don't render the warmup status
+		// indicator AND typically finish before LSP servers would have stabilized — warming them just spends
+		// CPU parsing big `initialize` responses concurrently with the LLM stream consumer, jittering
+		// perceived latency.
 		let lspServers: CreateAgentSessionResult["lspServers"];
-		if (enableLsp && options.hasUI && settings.get("lsp.diagnosticsOnWrite")) {
+		if (enableLsp && options.hasUI && settings.get("lsp.lazy")) {
+			lspServers = discoverStartupLspServers(cwd, "available");
+		} else if (enableLsp && options.hasUI) {
 			lspServers = discoverStartupLspServers(cwd);
 			if (lspServers.length > 0) {
 				void (async () => {

package/src/session/agent-session.ts

@@ -79,14 +79,14 @@ import {
 	clearAnthropicFastModeFallback,
 	deriveClaudeDeviceId,
 	Effort,
-	getSupportedEfforts,
 	isContextOverflow,
 	isUsageLimitError,
-	modelsAreEqual,
 	parseRateLimitReason,
 	resolveServiceTier,
 	streamSimple,
 } from "@oh-my-pi/pi-ai";
+import { getSupportedEfforts } from "@oh-my-pi/pi-catalog/model-thinking";
+import { modelsAreEqual } from "@oh-my-pi/pi-catalog/models";
 import { countTokens, MacOSPowerAssertion } from "@oh-my-pi/pi-natives";
 import {
 	extractRetryHint,
@@ -105,7 +105,7 @@ import { classifyDifficulty } from "../auto-thinking/classifier";
 import { reset as resetCapabilities } from "../capability";
 import type { Rule } from "../capability/rule";
 import { shouldEnableAppendOnlyContext } from "../config/append-only-context-mode";
-import { MODEL_ROLE_IDS, type ModelRegistry } from "../config/model-registry";
+import type { ModelRegistry } from "../config/model-registry";
 import {
 	extractExplicitThinkingSelector,
 	formatModelSelectorValue,
@@ -115,6 +115,7 @@ import {
 	type ResolvedModelRoleValue,
 	resolveModelRoleValue,
 } from "../config/model-resolver";
+import { MODEL_ROLE_IDS } from "../config/model-roles";
 import { expandPromptTemplate, type PromptTemplate } from "../config/prompt-templates";
 import type { Settings, SkillsSettings } from "../config/settings";
 import { onAppendOnlyModeChanged } from "../config/settings";
@@ -287,6 +288,20 @@ export type AgentSessionEventListener = (event: AgentSessionEvent) => void;
 export type AsyncJobSnapshotItem = Pick<AsyncJob, "id" | "type" | "status" | "label" | "startTime">;
 
 const EMPTY_STOP_MAX_RETRIES = 3;
+const RETRY_BACKOFF_MAX_DELAY_MS = 8_000;
+const RETRY_BACKOFF_JITTER_RATIO = 0.25;
+
+function calculateRetryBackoffDelayMs(baseDelayMs: number, attempt: number): number {
+	const cappedDelayMs = Math.min(Math.max(0, baseDelayMs) * 2 ** Math.max(0, attempt - 1), RETRY_BACKOFF_MAX_DELAY_MS);
+	const jitter = 1 - Math.random() * RETRY_BACKOFF_JITTER_RATIO;
+	return cappedDelayMs * jitter;
+}
+
+/**
+ * Slack added past a sibling credential's block expiry before retrying, so
+ * the next getApiKey lands after the block has actually lapsed.
+ */
+const SIBLING_UNBLOCK_BUFFER_MS = 1_000;
 const NON_WHITESPACE_RE = /\S/;
 
 function hasNonWhitespace(value: string): boolean {
@@ -8307,13 +8322,16 @@ export class AgentSession {
 
 		const errorMessage = message.errorMessage || "Unknown error";
 		const parsedRetryAfterMs = this.#parseRetryAfterMsFromError(errorMessage);
-		let delayMs = retrySettings.baseDelayMs * 2 ** (this.#retryAttempt - 1);
+		let delayMs = calculateRetryBackoffDelayMs(retrySettings.baseDelayMs, this.#retryAttempt);
 		let switchedCredential = false;
 		let switchedModel = false;
+		// Set when a usage-limit error pinned the wait to credential
+		// availability — suppresses the generic retry-after bump below.
+		let usageLimitWaitMs: number | undefined;
 
 		if (this.model && isUsageLimitError(errorMessage)) {
 			const retryAfterMs = parsedRetryAfterMs ?? calculateRateLimitBackoffMs(parseRateLimitReason(errorMessage));
-			const switched = await this.#modelRegistry.authStorage.markUsageLimitReached(
+			const outcome = await this.#modelRegistry.authStorage.markUsageLimitReached(
 				this.model.provider,
 				this.sessionId,
 				{
@@ -8321,12 +8339,28 @@ export class AgentSession {
 					baseUrl: this.model.baseUrl,
 				},
 			);
-			if (switched) {
+			if (outcome.switched) {
 				switchedCredential = true;
 				delayMs = 0;
-			} else if (retryAfterMs > delayMs) {
-				// No more accounts to switch to — wait out the backoff
-				delayMs = retryAfterMs;
+			} else {
+				// No sibling credential is usable right now. Wait for whichever
+				// comes first: the provider's retry-after window for the current
+				// account, or the earliest moment a temporarily blocked sibling
+				// frees up (e.g. a 60s post-401 block or a 5-min usage-probe
+				// block) — the next attempt's getApiKey re-ranks and picks it up.
+				// Without this, one short-lived sibling block escalates a
+				// recoverable situation into the provider's multi-hour wait and
+				// trips the fail-fast cap below.
+				usageLimitWaitMs = retryAfterMs;
+				if (outcome.retryAtMs !== undefined) {
+					const siblingWaitMs = Math.max(0, outcome.retryAtMs - Date.now()) + SIBLING_UNBLOCK_BUFFER_MS;
+					if (siblingWaitMs < usageLimitWaitMs) {
+						usageLimitWaitMs = siblingWaitMs;
+					}
+				}
+				if (usageLimitWaitMs > delayMs) {
+					delayMs = usageLimitWaitMs;
+				}
 			}
 		}
 
@@ -8338,7 +8372,7 @@ export class AgentSession {
 			}
 			if (switchedModel) {
 				delayMs = 0;
-			} else if (parsedRetryAfterMs && parsedRetryAfterMs > delayMs) {
+			} else if (usageLimitWaitMs === undefined && parsedRetryAfterMs && parsedRetryAfterMs > delayMs) {
 				delayMs = parsedRetryAfterMs;
 			}
 		}

package/src/session/auth-broker-config.ts

@@ -65,13 +65,42 @@ async function readConfigYaml(): Promise<ConfigSnapshot> {
 	}
 }
 
+/**
+ * Process-lifetime memo for {@link resolveAuthBrokerConfig}. Keyed on the env
+ * inputs (plus agent dir, which decides which config.yml is read) so tests
+ * that flip `OMP_AUTH_BROKER_*` between cases still observe the change, while
+ * repeated resolution within one CLI invocation (startup, subagent sessions)
+ * skips the config.yml read and any `!command` token resolution.
+ */
+let cachedConfigKey: string | null = null;
+let cachedConfigPromise: Promise<AuthBrokerClientConfig | null> | null = null;
+
 /**
  * Read broker configuration. Returns null when the URL is missing
  * (broker disabled — local store is used). Throws when URL is set but no
  * token is available — the caller cannot fall back silently because the
  * user explicitly asked to use the broker.
+ *
+ * Successful resolutions (including "no broker configured") are memoized for
+ * the process lifetime; failures are not, so a missing token can be fixed and
+ * retried. Concurrent callers share one in-flight resolution.
  */
-export async function resolveAuthBrokerConfig(): Promise<AuthBrokerClientConfig | null> {
+export function resolveAuthBrokerConfig(): Promise<AuthBrokerClientConfig | null> {
+	const key = `${process.env.OMP_AUTH_BROKER_URL ?? ""}\u0000${process.env.OMP_AUTH_BROKER_TOKEN ?? ""}\u0000${getAgentDir()}`;
+	if (cachedConfigPromise && cachedConfigKey === key) return cachedConfigPromise;
+	const promise = resolveAuthBrokerConfigUncached();
+	cachedConfigKey = key;
+	cachedConfigPromise = promise;
+	promise.catch(() => {
+		if (cachedConfigPromise === promise) {
+			cachedConfigPromise = null;
+			cachedConfigKey = null;
+		}
+	});
+	return promise;
+}
+
+async function resolveAuthBrokerConfigUncached(): Promise<AuthBrokerClientConfig | null> {
 	const envUrl = process.env.OMP_AUTH_BROKER_URL;
 	const envToken = process.env.OMP_AUTH_BROKER_TOKEN;
 

package/src/session/session-manager.ts

@@ -1516,10 +1516,10 @@ class NdjsonFileWriter {
 	}
 }
 
-/** Get recent sessions for display in welcome screen */
+/** Get recent sessions for display in welcome screen (which reserves WELCOME_SESSION_SLOTS rows) */
 export async function getRecentSessions(
 	sessionDir: string,
-	limit = 3,
+	limit = 4,
 	storage: SessionStorage = new FileSessionStorage(),
 ): Promise<RecentSessionInfo[]> {
 	const sessions = await getSortedSessions(sessionDir, storage);

package/src/session/streaming-output.ts

@@ -650,6 +650,7 @@ export class OutputSink {
 	#sawData = false;
 	#truncated = false;
 	#lastChunkTime = 0;
+	#pendingChunk = "";
 
 	// Per-line column cap streaming state (persists across `push` calls so a
 	// long line split across chunks still trips the same trigger).
@@ -701,14 +702,20 @@ export class OutputSink {
 	push(chunk: string): void {
 		chunk = sanitizeWithOptionalSixelPassthrough(chunk, sanitizeText);
 
-		// Throttled onChunk: only call the callback when enough time has passed.
+		// Throttled onChunk: coalesce chunks arriving inside the throttle window
+		// and flush the buffered concatenation on the next eligible tick (plus a
+		// final flush in dump()) so the preview never has silent gaps.
 		// Live preview gets the raw (pre-cap) chunk so the TUI never lags behind
 		// what reached the sink — the column cap is for the persisted LLM view.
 		if (this.#onChunk) {
 			const now = Date.now();
 			if (now - this.#lastChunkTime >= this.#chunkThrottleMs) {
 				this.#lastChunkTime = now;
-				this.#onChunk(chunk);
+				const merged = this.#pendingChunk + chunk;
+				this.#pendingChunk = "";
+				this.#onChunk(merged);
+			} else {
+				this.#pendingChunk += chunk;
 			}
 		}
 
@@ -880,6 +887,11 @@ export class OutputSink {
 			const sink = Bun.file(this.#artifactPath).writer();
 			this.#file = { path: this.#artifactPath, artifactId: this.#artifactId, sink };
 
+			// Head-retained bytes precede the rolling tail buffer in the capture.
+			if (this.#head.length > 0) {
+				sink.write(this.#head);
+			}
+
 			// Flush existing buffer to file BEFORE it gets trimmed further.
 			if (this.#buffer.length > 0) {
 				sink.write(this.#buffer);
@@ -946,10 +958,19 @@ export class OutputSink {
 		this.#columnEllipsisAdded = false;
 		this.#columnDroppedBytes = 0;
 		this.#columnTruncatedLines = 0;
+		this.#pendingChunk = "";
 	}
 
 	async dump(notice?: string): Promise<OutputSummary> {
 		const noticeLine = notice ? `[${notice}]\n` : "";
+
+		// Flush any chunk still held back by the throttle so the live preview
+		// ends with the complete stream.
+		if (this.#onChunk && this.#pendingChunk.length > 0) {
+			const pending = this.#pendingChunk;
+			this.#pendingChunk = "";
+			this.#onChunk(pending);
+		}
 		const totalLines = this.#sawData ? this.#totalLines + 1 : 0;
 
 		if (this.#file) await this.#file.sink.end();

package/src/slash-commands/builtin-registry.ts

@@ -30,6 +30,7 @@ import { createMarketplaceManager } from "./helpers/marketplace-manager";
 import { handleMcpAcp } from "./helpers/mcp";
 import { commandConsumed, errorMessage, parseSlashCommand, parseSubcommand, usage } from "./helpers/parse";
 import { handleSshAcp } from "./helpers/ssh";
+import { launchStatsDashboard, parseStatsDashboardArgs } from "./helpers/stats-dashboard";
 import { handleTodoAcp } from "./helpers/todo";
 import { buildUsageReportText } from "./helpers/usage-report";
 import { parseMarketplaceInstallArgs, parsePluginScopeArgs } from "./marketplace-install-parser";
@@ -542,6 +543,25 @@ const BUILTIN_SLASH_COMMAND_REGISTRY: ReadonlyArray<SlashCommandSpec> = [
 			runtime.ctx.editor.setText("");
 		},
 	},
+	{
+		name: "stats",
+		description: "Launch the local stats dashboard",
+		inlineHint: "[--port <port>]",
+		allowArgs: true,
+		handle: async (command, runtime) => {
+			const parsed = parseStatsDashboardArgs(command.args);
+			if ("error" in parsed) return usage(parsed.error, runtime);
+
+			await runtime.output("Syncing session files...");
+			try {
+				const result = await launchStatsDashboard(parsed);
+				await runtime.output(result.message);
+			} catch (error) {
+				await runtime.output(`Stats dashboard failed: ${errorMessage(error)}`);
+			}
+			return commandConsumed();
+		},
+	},
 	{
 		name: "changelog",
 		description: "Show changelog entries",