@ferris1225/pi-subagents 0.11.0 → 0.13.0

package/README.md

@@ -23,6 +23,10 @@ agent, and keep the workflow moving without manual polling.
   elapsed time; completion also produces a concise notification.
 - **Per-agent configuration** — enable agents, pick model and thinking strength per agent,
   tune concurrency limits, and choose discovery scope from `/subagents-setup`.
+- **Automatic model fallback** — if an agent's model fails at the provider level before
+  producing any output, the run is retried once with the main window's current model.
+  Per-run only, never persisted: a transient provider hiccup does not silently downgrade
+  the configured model.
 - **Leaf processes** — child agents cannot access the `subagent` tool, so delegation cannot
   recurse.
 
@@ -44,15 +48,198 @@ The default configuration enables `explore`, `worker`, and `reviewer`.
 
 ## Included agents
 
-| Agent | Default | Access | Purpose |
-| --- | :---: | --- | --- |
-| `explore` | Yes | Read-only | Fast codebase reconnaissance and structured findings. |
-| `worker` | Yes | Full | Implements, fixes, refactors, and tests a self-contained task. |
-| `reviewer` | Yes | Read-only | Independent adversarial review of a diff before completion. |
+| Agent | Default | Access | Default model | Thinking | Purpose |
+| --- | :---: | --- | --- | --- | --- |
+| `explore` | Yes | Read-only | `claude-haiku-4-5` | `low` | Fast codebase reconnaissance and structured findings. |
+| `worker` | Yes | Full | `claude-sonnet-4-5` | `high` | Implements, fixes, refactors, and tests a self-contained task. |
+| `reviewer` | Yes | Read-only | `claude-sonnet-4-5` | `high` | Adversarial quality gate: diff review (default), plus plan, proposed-solution, codebase-health, and PR/issue validation. |
 
 Agents are Markdown files in `agents/`. Each file contains YAML frontmatter and a system
-prompt. User and project scopes can override a built-in agent with the same name.
+prompt. User and project scopes can override a built-in agent with the same name; the
+frontmatter defaults above are overridden by `agentModels` / `agentThinkingLevels` when set.
+
+### Agent prompts
+
+The prompts below mirror `agents/*.md` — the source of truth loaded at dispatch time. They are
+the contract: each agent's role, hard constraints, and output format. Prompt drift shows up
+here first.
+
+<details>
+<summary><code>agents/explore.md</code> — reconnaissance</summary>
+
+```markdown
+---
+name: explore
+description: Fast read-only codebase reconnaissance. Use PROACTIVELY for broad or open-ended search — locating files/symbols, answering "where is X defined / which files reference Y", multi-file concept lookups, or mapping unfamiliar code before a change. Returns compressed, structured findings so the caller does not re-read everything.
+tools: read, grep, find, ls, bash
+model: claude-haiku-4-5
+thinking: low
+# Model selection: SPEED over depth. Pick the fastest available model.
+# What matters: fast grep/find/read, structured output. What doesn't: deep reasoning.
+---
+
+You are an explore agent: a fast, read-only reconnaissance specialist. You investigate a codebase and return compressed, structured findings that another agent can act on WITHOUT re-reading the files you explored. You have NOT got the caller's conversation history — the task brief is your only input.
+
+## Hard constraints
+- You are READ-ONLY. Never create, edit, or delete files; never run mutating commands.
+- Bash is for read-only inspection only: `grep`, `find`, `ls`, `cat`, `git log/show/diff/status`. No installs, builds, or state changes.
+- Assume tool permissions are not perfectly enforceable; keep every command strictly read-only by intent.
+
+## When invoked
+1. Orient with `grep`/`find` to locate the relevant code fast. Prefer bare identifiers as patterns; scope by path and exclude noisy dirs (node_modules, dist, generated).
+2. Read KEY SECTIONS, not whole files. After 1-2 greps, read the top match instead of running more greps.
+3. Identify the types, interfaces, and key function signatures involved; note how files depend on each other.
+4. Record exact paths and line ranges so the caller can jump straight in.
+
+## Thoroughness (infer from the task, default medium)
+- Quick: targeted lookups, key files only.
+- Medium: follow imports and callers, read critical sections.
+- Thorough: trace dependencies across modules; check tests and types.
+
+## Collaboration
+- Your output feeds `worker` (or the main agent directly). Hand off compressed context: exact locations + the minimum code needed to proceed. Flag anything ambiguous so the caller can decide.
+
+## Output format
+## Files Retrieved
+1. `path/to/file.ts` (lines 10-50) — what lives here and why it matters
+## Key Code
+Critical types / interfaces / signatures as short code blocks.
+## Architecture
+A brief explanation of how the pieces connect.
+## Start Here
+Which file to look at first, and why.
+
+## Quality standards
+Terse and factual. Exact paths and line numbers. Compress — do not narrate your search process or pad with prose.
+```
+
+</details>
+
+<details>
+<summary><code>agents/worker.md</code> — implementation</summary>
+
+```markdown
+---
+name: worker
+description: General-purpose implementation agent with full tools in an isolated context. Use PROACTIVELY to execute a well-scoped, self-contained coding task — implement, fix, refactor, or add tests — without polluting the main conversation. Plans internally, then implements and verifies. Give it a complete, self-contained brief.
+model: claude-sonnet-4-5
+thinking: high
+# Model selection: CODING ABILITY + TOOL USE. The primary implementation model —
+# balance quality against cost. No `tools` field => inherits all tools (full capability).
+---
+
+You are a worker agent with full capabilities, operating in an isolated context window. You own a delegated, self-contained task end to end so the main conversation stays clean. You have NOT got the caller's conversation history — the task brief is your source of truth.
+
+## Standard operating procedure
+Work in phases. Do not skip planning or verification.
+
+### Phase 1 — Context
+Read the brief fully. If it references files, read them before editing. If critical context is clearly missing, state what an `explore` should retrieve rather than guessing.
+
+### Phase 2 — Plan
+Inspect existing code and conventions first. Form the smallest coherent root-cause change that satisfies the brief. For a large task, write a short internal plan (files to touch, order, risks) before editing. Do not refactor unrelated code or create docs unless the brief asks.
+
+### Phase 3 — Implement
+Make the change. Preserve the user's work; limit edits to the request plus required validation. Follow the project's existing error handling, naming, and style.
 
+### Phase 4 — Verify
+Run the project's format/build/tests when they exist (e.g. `tsc --noEmit`, the test runner). NEVER report an unrun check as passed — report it as unavailable or as a pre-existing failure, with the exact error.
+
+### Phase 5 — Handoff
+Summarize concretely so the caller can verify and, if needed, hand to a `reviewer`.
+
+## Collaboration
+- You cannot dispatch sub-agents (children are leaf processes with no `subagent` tool). When the
+  brief lacks context that needs broad code discovery, state concretely what an `explore` should
+  retrieve for the caller — do not guess.
+- Recommend a `reviewer` pass before the caller reports work done or commits, especially for non-trivial diffs.
+
+## Output format
+## Completed
+What was done, in a few lines.
+## Files Changed
+- `path/to/file.ts` — what changed.
+## Verification
+Which checks you ACTUALLY ran and their result (e.g. `tsc --noEmit` clean; `vitest` 12 passed). State explicitly anything you could not run and why.
+## Notes (if any)
+Follow-ups, decisions made, blockers. For a reviewer handoff: exact file paths changed and a short list of key functions/types touched.
+
+## Quality standards
+Root-cause fixes over patches. No unrelated churn. Honest verification — an unrun check is never a passed check.
+```
+
+</details>
+
+<details>
+<summary><code>agents/reviewer.md</code> — quality gate</summary>
+
+```markdown
+---
+name: reviewer
+description: Adversarial code reviewer and pre-commit quality gate. Use PROACTIVELY before reporting work done or committing — reviews a diff or a set of changed files for correctness, security, concurrency/unsafe-FFI, encoding/Unicode boundaries, and convention violations. Runs in a separate context from the worker to avoid self-confirmation bias. Read-only; never edits, builds, or runs tests. Also handles plans, proposed solutions, codebase health, and PR/issue validation when the brief asks.
+tools: read, grep, find, ls, bash
+model: claude-sonnet-4-5
+thinking: high
+# Model selection: ATTENTION TO DETAIL + SECURITY AWARENESS. This is the quality gate —
+# use the strongest available reasoning model.
+---
+
+You are a senior, adversarial code reviewer. Your job is to FIND WHAT IS WRONG, not to validate. Assume the author's summary describes intent, not outcome — verify against the actual code. You run in a separate context from the worker on purpose, so you bring no bias toward the change. You have NOT got the caller's conversation history.
+
+## Hard constraints
+- You are READ-ONLY. Do NOT modify files, run builds, or run tests.
+- Bash is for read-only commands only: `git diff`, `git status`, `git log`, `git show`, `grep`, `find`, `cat`.
+- Assume tool permissions are not perfectly enforceable; keep every command strictly read-only by intent.
+
+## Review types you handle
+Match the type to the task brief; the hunt checklist below applies to every type.
+
+### 1. Code diffs (default)
+1. Run `git diff` and `git status` to see the recent changes. If a specific file set was given, read those files.
+2. Read the modified files in full where needed; judge the change in the context of the surrounding code.
+
+### 2. Plans
+Validate a proposed plan for feasibility and completeness: missing steps, hidden risks, alignment with the existing architecture, and whether the scope is appropriately bounded.
+
+### 3. Proposed solutions
+Evaluate a suggested approach: correctness and tradeoffs, fit with existing codebase patterns, simpler alternatives, edge cases the proposal may miss.
+
+### 4. Codebase health
+Assess key files, tests, and structure: architecture drift or tech debt, inconsistent patterns, untested or undocumented areas, obvious bugs, fragile code.
+
+### 5. Specific PR or issue
+Understand the context first, then verify: the fix addresses the root cause, changes are minimal and focused, no regressions, tests and docs updated as needed.
+
+## Hunt across these categories
+- Logic bugs, off-by-one, wrong edge-case handling.
+- Error handling gaps; swallowed failures; unreported unrun checks.
+- Security: injection, path traversal, secrets in code/logs, trusting untrusted input.
+- Concurrency: shared mutable state, locks held across await, races.
+- Encoding/Unicode: assuming `char*`/files/CLI text is UTF-8; wrong `A` vs `W` Win32 APIs; boundary conversions.
+- Resource leaks; violations of the project's stated conventions.
+- Classify severity honestly. Distinguish blockers from nits; do not pad with style preferences.
+
+## Collaboration
+- Independent of `worker` by design — your verdict is the gate before commit. Fix nothing yourself; report so the caller can dispatch a worker.
+
+## Output format
+## Files Reviewed
+- `path/to/file.ts`
+## Critical (must fix)
+- `file.ts:42` — concrete issue and why it breaks.
+## Warnings (should fix)
+- `file.ts:10` — issue and suggested direction.
+## Suggestions (consider)
+- Optional improvements.
+## Verdict
+One of: APPROVE / APPROVE_WITH_NITS / REQUEST_CHANGES, plus a 2-3 sentence rationale.
+End with exactly one machine-readable line: `VERDICT: REVIEW_PASS` for APPROVE or APPROVE_WITH_NITS; `VERDICT: REVIEW_FAIL` for REQUEST_CHANGES.
+
+## Quality standards
+Specific file paths and line numbers. No vague feedback. A clean report means you looked hard, not that you found nothing to say.
+```
+
+</details>
 ## Workflow
 
 A typical flow is:
@@ -70,9 +257,9 @@ main agent
 1. The main agent calls `subagent` with a self-contained brief.
 2. The tool returns immediately and ends that foreground tool turn, leaving the editor ready
    for input.
-3. The child process works independently. By default up to four queued runs execute at
-   once and a single parallel request may contain up to eight tasks; both limits are
-   configurable (`maxConcurrency` / `maxParallelTasks`).
+3. The child process works independently. By default up to four sub-agents run at once —
+   and one parallel call accepts at most four tasks; extra runs queue up to `maxConcurrency`
+   (configurable via `/subagents-setup` or `pi-subagents.json`).
 4. On completion or failure, the extension sends a durable result message to the main
    session. That message automatically wakes the main agent, or waits until its current turn
    finishes.
@@ -144,8 +331,6 @@ agent's default — its frontmatter `thinking`, else the global default). The gl
   "proactiveInjection": true,
   "agentScope": "user",
   "maxConcurrency": 4,
-  "maxParallelTasks": 8,
-  "maxSubagentDepth": 1,
   "maxFixRounds": 2
 }
 ```
@@ -160,9 +345,7 @@ agent's default — its frontmatter `thinking`, else the global default). The gl
 | `maxResultLines` | Max lines of a sub-agent result carried in the completion message (default `80`). Longer results are truncated; the full text is written to a temp file whose path is included in the message. |
 | `proactiveInjection` | Whether to add the delegation directive to the main system prompt. |
 | `agentScope` | `user`, `project`, or `both`; controls which user/project agent directories are discovered. |
-| `maxConcurrency` | How many sub-agent processes run at once (1–16, default 4). Extra work waits in the queue. |
-| `maxParallelTasks` | Maximum tasks accepted by one parallel `subagent` call (1–32, default 8). |
-| `maxSubagentDepth` | Depth at which the `subagent` tool is no longer registered (default 1: the main session delegates, children are leaf processes). `0` disables the tool entirely. Read once at extension load. |
+| `maxConcurrency` | Max sub-agent processes running at once (1–16, default 4), and the max tasks one parallel `subagent` call accepts. Extra work waits in the queue. |
 | `maxFixRounds` | Auto-fix rounds when a reviewer returns `REVIEW_FAIL`: the extension dispatches a `worker` (briefed with the review's concrete findings) then a `reviewer` re-review, repeating up to this many times before waking the main agent with the full chain. `0` disables it (the main agent handles fixes itself). Default 2. The reviewer stays read-only and in its own context; the loop is orchestrated by the extension, not by the reviewer. |
 
 ### Configuration migration
@@ -173,6 +356,11 @@ The config file migrates itself on load — no manual steps after an upgrade:
   holding invalid values) is normalized and saved back with the new fields filled in.
 - **Removed agents** — agents no longer shipped (e.g. the old `plan` agent) are stripped
   from `enabledAgents`, `agentModels`, and `agentThinkingLevels` automatically.
+- **Merged limits** — the pre-0.13 `maxParallelTasks` key is folded into `maxConcurrency`
+  (the larger of the two wins) and dropped on the next save.
+- **Removed keys** — `maxSubagentDepth` (0.14) is dropped on load: sub-agent children are
+  always leaf processes (the `subagent` tool is excluded from their toolset, with a depth
+  marker as defense in depth). To disable delegation entirely, use `"enabledAgents": []`.
 
 Model selection uses this precedence:
 
@@ -183,6 +371,13 @@ configured agent model → current main-session model → agent frontmatter mode
 Unavailable configured models are replaced with a usable current-session model when possible
 and the repaired configuration is saved.
 
+At runtime, if an agent's model fails at the provider level before producing any output (bad
+model id, auth, thinking level, quota, ...), the run is retried **once** with the main window's
+current model. This per-run degradation is never persisted — a transient provider hiccup must
+not silently downgrade the configured model — and it does not apply to task-level failures
+(the model worked, the task failed), aborts, or timeouts. Results carry a `model fell back
+from …` note when it happened.
+
 Thinking strength uses this precedence: `agentThinkingLevels` entry → agent frontmatter `thinking` → `thinkingLevel` default.
 
 ## Agent discovery and overrides

package/agents/worker.md

@@ -28,7 +28,9 @@ Run the project's format/build/tests when they exist (e.g. `tsc --noEmit`, the t
 Summarize concretely so the caller can verify and, if needed, hand to a `reviewer`.
 
 ## Collaboration
-- Request `explore` first when the task needs broad code discovery you were not given.
+- You cannot dispatch sub-agents (children are leaf processes with no `subagent` tool). When the
+  brief lacks context that needs broad code discovery, state concretely what an `explore` should
+  retrieve for the caller — do not guess.
 - Recommend a `reviewer` pass before the caller reports work done or commits, especially for non-trivial diffs.
 
 ## Output format

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@ferris1225/pi-subagents",
-  "version": "0.11.0",
+  "version": "0.13.0",
   "description": "Focused sub-agent delegation for pi: explore / worker / reviewer agents in isolated context, with proactive dispatch injection and per-agent model selection.",
   "type": "module",
   "license": "MIT",

package/src/config.ts

@@ -45,21 +45,10 @@ export const MAX_RESULT_LINES_LIMIT = 2000;
 
 export const CONFIG_FILE_NAME = "pi-subagents.json";
 
-/** How many sub-agent processes may run at once. Default: 4. */
+/** How many sub-agent processes may run at once, and how many tasks one parallel `subagent` call may contain. Default: 4. */
 export const DEFAULT_MAX_CONCURRENCY = 4;
 /** Upper bound accepted for maxConcurrency (defensive clamp). */
 export const MAX_CONCURRENCY_LIMIT = 16;
-/** How many tasks a single parallel `subagent` call may contain. Default: 8. */
-export const DEFAULT_MAX_PARALLEL_TASKS = 8;
-/** Upper bound accepted for maxParallelTasks (defensive clamp). */
-export const MAX_PARALLEL_TASKS_LIMIT = 32;
-/**
- * Depth at which the subagent tool stops being available. 1 = the main session
- * delegates and child processes are leaves; 0 disables the tool entirely.
- */
-export const DEFAULT_MAX_SUBAGENT_DEPTH = 1;
-/** Upper bound accepted for maxSubagentDepth (defensive clamp). */
-export const MAX_SUBAGENT_DEPTH_LIMIT = 4;
 /**
  * How many automatic worker→reviewer fix rounds run when a reviewer returns
  * REVIEW_FAIL before waking the main agent. 0 disables the auto-fix loop
@@ -93,12 +82,9 @@ export interface SubagentsConfig {
 	proactiveInjection: boolean;
 	/** Which agent directories to discover from. Default: "user". */
 	agentScope: AgentScope;
-	/** Max sub-agent processes running at once (extra work queues). Default: 4. */
+	/** Max sub-agent processes running at once (extra work queues) and the max tasks
+	 * one parallel `subagent` call may contain. Default: 4. */
 	maxConcurrency: number;
-	/** Max tasks accepted by one parallel `subagent` call. Default: 8. */
-	maxParallelTasks: number;
-	/** Depth at which the subagent tool is no longer registered. Default: 1. */
-	maxSubagentDepth: number;
 	/**
 	 * Auto-fix rounds when a reviewer returns REVIEW_FAIL: the extension dispatches
 	 * a worker (briefed with the review's concrete findings) then a reviewer
@@ -119,8 +105,6 @@ export const DEFAULT_CONFIG: SubagentsConfig = {
 	proactiveInjection: true,
 	agentScope: "user",
 	maxConcurrency: DEFAULT_MAX_CONCURRENCY,
-	maxParallelTasks: DEFAULT_MAX_PARALLEL_TASKS,
-	maxSubagentDepth: DEFAULT_MAX_SUBAGENT_DEPTH,
 	maxFixRounds: DEFAULT_MAX_FIX_ROUNDS,
 };
 
@@ -166,8 +150,6 @@ export function normalizeConfig(raw: unknown): SubagentsConfig {
 		proactiveInjection: DEFAULT_CONFIG.proactiveInjection,
 		agentScope: DEFAULT_CONFIG.agentScope,
 		maxConcurrency: DEFAULT_CONFIG.maxConcurrency,
-		maxParallelTasks: DEFAULT_CONFIG.maxParallelTasks,
-		maxSubagentDepth: DEFAULT_CONFIG.maxSubagentDepth,
 		maxFixRounds: DEFAULT_CONFIG.maxFixRounds,
 	};
 
@@ -223,12 +205,12 @@ export function normalizeConfig(raw: unknown): SubagentsConfig {
 	const maxConcurrency = clampCount(raw.maxConcurrency, MAX_CONCURRENCY_LIMIT);
 	if (maxConcurrency !== undefined) config.maxConcurrency = maxConcurrency;
 
-	const maxParallelTasks = clampCount(raw.maxParallelTasks, MAX_PARALLEL_TASKS_LIMIT);
-	if (maxParallelTasks !== undefined) config.maxParallelTasks = maxParallelTasks;
-
-	// 0 is meaningful here (disables the tool), so clamp to [0, limit] instead.
-	if (typeof raw.maxSubagentDepth === "number" && Number.isFinite(raw.maxSubagentDepth)) {
-		config.maxSubagentDepth = Math.max(0, Math.min(MAX_SUBAGENT_DEPTH_LIMIT, Math.round(raw.maxSubagentDepth)));
+	// Schema migration: maxParallelTasks (pre-0.13) merged into maxConcurrency.
+	// Take the larger of the two so an upgraded config never loses capacity it
+	// was explicitly given; the old key is dropped on the persisted save.
+	const legacyParallelTasks = clampCount(raw.maxParallelTasks, MAX_CONCURRENCY_LIMIT);
+	if (legacyParallelTasks !== undefined && legacyParallelTasks > config.maxConcurrency) {
+		config.maxConcurrency = legacyParallelTasks;
 	}
 
 	// 0 disables the auto-fix loop (main agent handles fixes itself).

package/src/index.ts

@@ -35,7 +35,7 @@ import {
 	getResultOutput,
 	isFailedResult,
 	reviewVerdict,
-	runSingleAgent,
+	runSingleAgentWithModelFallback,
 	truncateResultOutput,
 	writeResultArtifact,
 	type SingleResult,
@@ -131,7 +131,10 @@ function formatCompletionBlock(result: SingleResult, maxResultLines: number): st
 	const usage = formatUsage(result.usage);
 	const output = getResultOutput(result);
 	const { text, truncated } = truncateResultOutput(output, maxResultLines);
-	const lines = [`### [${result.agent}] ${status}${usage ? ` (${usage})` : ""}`, "", `Task: ${formatTaskSummary(result.task)}`, "", text];
+	const fallbackNote = result.modelFallbackFrom
+		? ` (model fell back from ${result.modelFallbackFrom} to ${result.model ?? "main-window model"})`
+		: "";
+	const lines = [`### [${result.agent}] ${status}${usage ? ` (${usage})` : ""}${fallbackNote}`, "", `Task: ${formatTaskSummary(result.task)}`, "", text];
 	if (truncated) {
 		// The full text lives on disk so the main agent can read it on demand.
 		lines.push("", `(output truncated to ${maxResultLines} lines; full result: ${writeResultArtifact(output, result.agent)})`);
@@ -164,17 +167,15 @@ export default function (pi: ExtensionAPI): void {
 	};
 	const completionBatcher = createCompletionBatcher<CompletionMessageItem>({ emit: sendCompletionGroup });
 
-	// Recursion guard: sub-agents at the configured depth are leaf processes and
-	// cannot delegate again. maxSubagentDepth 0 disables the tool entirely.
-	if (currentSubagentDepth() >= initialConfig.maxSubagentDepth) {
-		const reason =
-			initialConfig.maxSubagentDepth === 0
-				? "disabled by maxSubagentDepth 0 in pi-subagents.json"
-				: "disabled in nested sub-agent processes";
+	// Recursion guard: sub-agent children are leaf processes. The `subagent` tool is
+	// excluded from their toolset at spawn (--exclude-tools); this check is defense
+	// in depth so a child can never expose the tool back to its model, even if
+	// another extension ignores the depth marker.
+	if (currentSubagentDepth() >= 1) {
 		pi.registerCommand("subagents-setup", {
-			description: `Configure pi-subagents (${reason})`,
+			description: "Configure pi-subagents (unavailable in nested sub-agent processes)",
 			handler: async (_args, ctx) => {
-				ctx.ui.notify(`pi-subagents setup is unavailable here (${reason}).`, "warning");
+				ctx.ui.notify("pi-subagents setup is unavailable in nested sub-agent processes.", "warning");
 			},
 		});
 		return;
@@ -352,16 +353,19 @@ export default function (pi: ExtensionAPI): void {
 				const runId = monitor.addRun(agent.name, task, agent.model, thinkingLevel, meta);
 				const onLive = makeLiveHandler(runId);
 				try {
-					const result = await runSingleAgent({
-						defaultCwd: ctx.cwd,
-						agent,
-						agentName,
-						task,
-						thinkingLevel,
-						signal,
-						onLive,
-						makeDetails: makeDetails("single", true),
-					});
+					const result = await runSingleAgentWithModelFallback(
+						{
+							defaultCwd: ctx.cwd,
+							agent,
+							agentName,
+							task,
+							thinkingLevel,
+							signal,
+							onLive,
+							makeDetails: makeDetails("single", true),
+						},
+						sessionRef,
+					);
 					finishRun(runId, isFailedResult(result) ? "failed" : "done");
 					return result;
 				} catch (error) {
@@ -438,17 +442,20 @@ export default function (pi: ExtensionAPI): void {
 					async (backgroundSignal) => {
 						let result: SingleResult;
 						try {
-							result = await runSingleAgent({
-								defaultCwd: ctx.cwd,
-								agent,
-								agentName,
-								task,
-								cwd,
-								thinkingLevel,
-								signal: backgroundSignal,
-								onLive,
-								makeDetails: makeDetails("single", true),
-							});
+							result = await runSingleAgentWithModelFallback(
+								{
+									defaultCwd: ctx.cwd,
+									agent,
+									agentName,
+									task,
+									cwd,
+									thinkingLevel,
+									signal: backgroundSignal,
+									onLive,
+									makeDetails: makeDetails("single", true),
+								},
+								sessionRef,
+							);
 						} catch (error) {
 							const errorMessage = error instanceof Error ? error.message : String(error);
 							result = {
@@ -495,12 +502,12 @@ export default function (pi: ExtensionAPI): void {
 			// Sub-agents intentionally detach from the foreground turn. This makes the
 			// editor available immediately; completion messages later wake the main agent.
 			if (params.tasks && params.tasks.length > 0) {
-				if (params.tasks.length > config.maxParallelTasks) {
+				if (params.tasks.length > config.maxConcurrency) {
 					return {
 						content: [
 							{
 								type: "text",
-								text: `Too many parallel tasks (${params.tasks.length}). Max is ${config.maxParallelTasks} (configurable via /subagents-setup).`,
+								text: `Too many parallel tasks (${params.tasks.length}). Max is ${config.maxConcurrency} (configurable via /subagents-setup).`,
 							},
 						],
 						details: makeDetails("parallel", true)([]),
@@ -570,7 +577,7 @@ export default function (pi: ExtensionAPI): void {
 				const pending = r.exitCode === -1;
 				const icon = statusIcon(pending ? "running" : isFailedResult(r) ? "failed" : "done", theme);
 				const usage = formatUsage(r.usage);
-				const model = r.model ?? "?";
+				const model = `${r.model ?? "?"}${r.modelFallbackFrom ? ` (fell back from ${r.modelFallbackFrom})` : ""}`;
 				const line = `${theme.fg("toolTitle", theme.bold("subagent "))}${icon} ${theme.fg("accent", r.agent)} ${theme.fg("dim", `· ${model}${r.thinking ? ` · thinking ${r.thinking}` : ""}${pending ? " · background" : ""}${usage ? ` · ${usage}` : ""}`)}`;
 				return new Text(line, 0, 0);
 			}
@@ -583,7 +590,7 @@ export default function (pi: ExtensionAPI): void {
 				const pending = r.exitCode === -1;
 				const icon = statusIcon(pending ? "running" : isFailedResult(r) ? "failed" : "done", theme);
 				const usage = formatUsage(r.usage);
-				const model = r.model ?? "?";
+				const model = `${r.model ?? "?"}${r.modelFallbackFrom ? ` (fell back from ${r.modelFallbackFrom})` : ""}`;
 				lines.push(`  ${icon} ${theme.fg("accent", r.agent)} ${theme.fg("dim", `· ${model}${r.thinking ? ` · thinking ${r.thinking}` : ""}${pending ? " · background" : ""}${usage ? ` · ${usage}` : ""}`)}`);
 			}
 			return new Text(lines.join("\n"), 0, 0);

package/src/setup.ts

@@ -16,7 +16,6 @@ import {
 	DEFAULT_ENABLED_AGENTS,
 	DEFAULT_MAX_CONCURRENCY,
 	DEFAULT_MAX_FIX_ROUNDS,
-	DEFAULT_MAX_PARALLEL_TASKS,
 	THINKING_LEVEL_VALUES,
 	type AgentScope,
 	type SubagentsConfig,
@@ -196,7 +195,6 @@ async function pickInjection(ctx: ExtensionCommandContext, current: boolean): Pr
 
 /** Preset steps offered for the two numeric limits (selection-only wizard). */
 const CONCURRENCY_STEPS = [1, 2, 3, 4, 6, 8, 12, 16];
-const PARALLEL_TASK_STEPS = [2, 4, 6, 8, 12, 16, 24, 32];
 /** Preset rounds offered for the auto-fix loop (0 disables it). */
 const FIX_ROUNDS_STEPS = [0, 1, 2, 3, 5];
 
@@ -284,22 +282,13 @@ async function runFullSetup(ctx: ExtensionCommandContext, configPath: string, ba
 
 	const maxConcurrency = await pickCount(
 		ctx,
-		"Max sub-agents running at once? (extra work queues)",
+		"Max sub-agents running at once (and per parallel call)? (extra work queues)",
 		CONCURRENCY_STEPS,
 		base.maxConcurrency,
 		DEFAULT_MAX_CONCURRENCY,
 	);
 	if (maxConcurrency === undefined) return notifyCancelled(ctx);
 
-	const maxParallelTasks = await pickCount(
-		ctx,
-		"Max tasks in one parallel subagent call?",
-		PARALLEL_TASK_STEPS,
-		base.maxParallelTasks,
-		DEFAULT_MAX_PARALLEL_TASKS,
-	);
-		if (maxParallelTasks === undefined) return notifyCancelled(ctx);
-
 		const maxFixRounds = await pickCount(
 			ctx,
 			"Auto-fix rounds when a reviewer returns REQUEST_CHANGES? (0 = main agent handles fixes)",
@@ -319,8 +308,6 @@ async function runFullSetup(ctx: ExtensionCommandContext, configPath: string, ba
 		proactiveInjection: injection,
 		agentScope: scope,
 		maxConcurrency,
-		maxParallelTasks,
-		maxSubagentDepth: base.maxSubagentDepth,
 		maxFixRounds,
 	};
 	await saveConfig(next, configPath);
@@ -335,7 +322,6 @@ async function runMenu(ctx: ExtensionCommandContext, configPath: string, config:
 		"Toggle proactive injection",
 		"Change agent scope",
 		"Change max concurrent sub-agents",
-		"Change max parallel tasks",
 		"Change max fix rounds",
 		"Full re-setup",
 	]);
@@ -381,23 +367,13 @@ async function runMenu(ctx: ExtensionCommandContext, configPath: string, config:
 	} else if (choice.startsWith("Change max concurrent")) {
 		const maxConcurrency = await pickCount(
 			ctx,
-			"Max sub-agents running at once? (extra work queues)",
+			"Max sub-agents running at once (and per parallel call)? (extra work queues)",
 			CONCURRENCY_STEPS,
 			config.maxConcurrency,
 			DEFAULT_MAX_CONCURRENCY,
 		);
 		if (maxConcurrency === undefined) return notifyCancelled(ctx);
 		next.maxConcurrency = maxConcurrency;
-	} else if (choice.startsWith("Change max parallel")) {
-		const maxParallelTasks = await pickCount(
-			ctx,
-			"Max tasks in one parallel subagent call?",
-			PARALLEL_TASK_STEPS,
-			config.maxParallelTasks,
-			DEFAULT_MAX_PARALLEL_TASKS,
-		);
-		if (maxParallelTasks === undefined) return notifyCancelled(ctx);
-		next.maxParallelTasks = maxParallelTasks;
 	} else if (choice.startsWith("Change max fix")) {
 		const maxFixRounds = await pickCount(
 			ctx,

package/src/spawn.ts

@@ -21,9 +21,8 @@ import type { AgentConfig, AgentSource } from "./agents.ts";
 import { DEFAULT_THINKING_LEVEL, type ThinkingLevel } from "./config.ts";
 
 /**
- * Limits are configurable: see maxConcurrency / maxParallelTasks /
- * maxSubagentDepth in config.ts (defaults 4 / 8 / 1, via /subagents-setup
- * or pi-subagents.json).
+ * Limits are configurable: see maxConcurrency in config.ts (default 4, via
+ * /subagents-setup or pi-subagents.json).
  */
 /** Default thinking level for sub-agents. pi clamps it to the resolved model's support. */
 export const SUBAGENT_THINKING_LEVEL: ThinkingLevel = DEFAULT_THINKING_LEVEL;
@@ -55,6 +54,8 @@ export interface SingleResult {
 	thinking?: string;
 	stopReason?: string;
 	errorMessage?: string;
+	/** Model the run degraded from: set when a failed run was retried with the main-window model. */
+	modelFallbackFrom?: string;
 }
 
 export interface SubagentDetails {
@@ -142,6 +143,23 @@ export function isFailedResult(result: SingleResult): boolean {
 	return result.exitCode !== 0 || result.stopReason === "error" || result.stopReason === "aborted";
 }
 
+/**
+ * True when a failed run never got usable output from its model: the provider
+ * rejected the call before the model produced any text (bad model id, auth,
+ * thinking level, quota, ...). Task-level failures — the model worked and the
+ * task failed — and aborts/timeouts are NOT model-level and must not degrade.
+ */
+export function isModelLevelFailure(result: SingleResult): boolean {
+	if (!isFailedResult(result)) return false;
+	if (result.stopReason === "aborted") return false;
+	// The model produced text: the failure belongs to the task, not the model.
+	if (getFinalOutput(result.messages)) return false;
+	if (result.errorMessage?.includes("timed out")) return false;
+	// Require evidence the failure came from the model/provider (an error
+	// message or stderr), not from the child process failing to start.
+	return result.messages.length > 0 || result.stderr.trim().length > 0;
+}
+
 export function getResultOutput(result: SingleResult): string {
 	if (isFailedResult(result)) {
 		return result.errorMessage || result.stderr || getFinalOutput(result.messages) || "(no output)";
@@ -525,3 +543,24 @@ export async function runSingleAgent(options: RunSingleOptions): Promise<SingleR
 			}
 	}
 }
+
+/**
+ * Run one agent; when the configured model fails at the provider level before
+ * producing any output (see isModelLevelFailure), retry once with the main
+ * window's current model. The retried result is returned with `modelFallbackFrom`
+ * set so callers can surface the degradation. The fallback is per-run only and
+ * never persisted: a transient provider hiccup must not silently downgrade the
+ * configured agent model.
+ */
+export async function runSingleAgentWithModelFallback(
+	options: RunSingleOptions,
+	fallbackModelRef?: string,
+): Promise<SingleResult> {
+	const result = await runSingleAgent(options);
+	const agent = options.agent;
+	const launchedRef = agent?.model;
+	if (!agent || !launchedRef || !fallbackModelRef || launchedRef === fallbackModelRef) return result;
+	if (!isModelLevelFailure(result)) return result;
+	const retried = await runSingleAgent({ ...options, agent: { ...agent, model: fallbackModelRef } });
+	return { ...retried, modelFallbackFrom: launchedRef };
+}