refactor(agent-loop): unify tool-call scheduling on one rolling pool + factory cap default

Run every ordered group through the same rolling pool: an exclusive call is a
pool of one (a barrier), dropping the separate runExclusive path and the
redundant post-grouping executionMode re-query. Behavior is unchanged — the
parallel-tool-calls snapshot and the full scheduler unit suite (barriers, cap,
abort, model-order results) stay green.

Add AgentLoop.Config.maxParallelToolCalls as a factory-wide default applied to
every agent create/createAgent/resume mints (per-agent option overrides it),
forwarded through agent-core so it reaches front doors that expose no cap field
of their own. Trim the isConcurrencySafe JSDoc to the local contract and link
the parallel-tool-call RFC for the full rationale; document the field on the
canonical core-data-structures page.

packages/core/agent-loop/src/index.ts

@@ -344,6 +344,16 @@ export { DEFAULT_MAX_PARALLEL_TOOL_CALLS } from './constants.ts'
 
 /** Plugin configuration for declarative startup agents. */
 export interface Config {
+  /**
+   * Default concurrent tool-call cap applied to every agent this factory
+   * creates (declarative startup agents and factory callers such as the ACP,
+   * stdio, and SDK front doors that go through `create`/`createAgent`/`resume`).
+   * A positive integer; a per-agent `maxParallelToolCalls` overrides it, and an
+   * agent with neither falls back to {@link DEFAULT_MAX_PARALLEL_TOOL_CALLS}.
+   * This is the single `cordis.yml` knob that reaches agents whose front door
+   * does not expose its own cap field.
+   */
+  maxParallelToolCalls?: number
   /** Agents created or resumed at plugin startup. */
   agents: (AgentOptions & {
     /** Registry identity for the live agent. */
@@ -366,6 +376,9 @@ export class AgentLoop extends Service implements AgentFactory {
 
   /** Runtime schema for declarative agents. */
   static Config = z.object({
+    // The factory-wide default cap; a per-agent value overrides it. A positive
+    // integer, validated here so a bad cordis.yml value fails at load.
+    maxParallelToolCalls: z.number().step(1).min(1),
     agents: z.array(z.object({
       id: z.string().required(),
       model: z.string(),
@@ -410,6 +423,22 @@ export class AgentLoop extends Service implements AgentFactory {
     }
   }
 
+  /**
+   * Merge the factory-wide default cap into one agent's options. A per-agent
+   * `maxParallelToolCalls` wins; otherwise the `Config.maxParallelToolCalls`
+   * default applies, reaching factory callers (ACP/stdio/SDK front doors) whose
+   * own config does not set a cap. Absent both, the loop falls back to
+   * {@link DEFAULT_MAX_PARALLEL_TOOL_CALLS} at schedule time.
+   * @param options - the caller-supplied agent options.
+   * @returns options with the default cap applied when the caller omitted one.
+   */
+  private withFactoryDefaults(options: AgentOptions): AgentOptions {
+    if (options.maxParallelToolCalls !== undefined || this.config.maxParallelToolCalls === undefined) {
+      return options
+    }
+    return { ...options, maxParallelToolCalls: this.config.maxParallelToolCalls }
+  }
+
   /**
    * Create an agent on a fresh per-run session, owned by the accessing fiber.
    * Constructor-driven config calls use the loop fiber itself.
@@ -419,13 +448,14 @@ export class AgentLoop extends Service implements AgentFactory {
    * @returns the published running agent.
    */
   create(id: AgentId, options: AgentOptions = {}, meta: Pick<SessionHeader, 'cwd'> = {}): ReactLoopAgent {
-    validateAgentOptions(options)
+    const resolved = this.withFactoryDefaults(options)
+    validateAgentOptions(resolved)
     const loopCtx = this.runtime.ctx
     const transaction = new AgentCreationTransaction(loopCtx, this.ctx, this.ownership, id)
     try {
       const sessionId = SessionId(`${id}-session-${randomUUID()}`)
       const session = loopCtx.sessions.prepare(sessionId, { meta })
-      const agent = transaction.prepare(options, session)
+      const agent = transaction.prepare(resolved, session)
       transaction.publish('startup')
       return agent
     } catch (error: unknown) {
@@ -443,7 +473,8 @@ export class AgentLoop extends Service implements AgentFactory {
    * @returns the published handle.
    */
   async createAgent(ownerCtx: Context, options: CreateAgentOptions): Promise<AgentHandle> {
-    validateAgentOptions(options.agentOptions ?? {})
+    const agentOptions = this.withFactoryDefaults(options.agentOptions ?? {})
+    validateAgentOptions(agentOptions)
     const transaction = new AgentCreationTransaction(
       this.runtime.ctx,
       ownerCtx,
@@ -456,7 +487,7 @@ export class AgentLoop extends Service implements AgentFactory {
         ...options.seed === undefined ? {} : { seed: options.seed },
         ...options.meta === undefined ? {} : { meta: options.meta },
       })
-      const agent = transaction.prepare(options.agentOptions ?? {}, session)
+      const agent = transaction.prepare(agentOptions, session)
       await transaction.waitFor(options.setup?.(agent.ctx))
       transaction.assertActive()
       return transaction.publish('startup')
@@ -475,7 +506,6 @@ export class AgentLoop extends Service implements AgentFactory {
    * @returns the published handle.
    */
   async resume(ownerCtx: Context, options: ResumeAgentOptions): Promise<AgentHandle> {
-    validateAgentOptions(options.agentOptions ?? {})
     const persistence = this.runtime.ctx.get('sessionPersistence')
     if (persistence === undefined) {
       throw new Error('cannot resume: session persistence is not configured (load a dsh-session-persistence backend)')
@@ -489,6 +519,8 @@ export class AgentLoop extends Service implements AgentFactory {
     persistence: SessionPersistence,
     options: ResumeAgentOptions,
   ): Promise<AgentHandle> {
+    const agentOptions = this.withFactoryDefaults(options.agentOptions ?? {})
+    validateAgentOptions(agentOptions)
     const transaction = new AgentCreationTransaction(
       this.runtime.ctx,
       ownerCtx,
@@ -508,7 +540,7 @@ export class AgentLoop extends Service implements AgentFactory {
           ...loaded.meta.seedLength === undefined ? {} : { seedLength: loaded.meta.seedLength },
         },
       })
-      const agent = transaction.prepare(options.agentOptions ?? {}, session)
+      const agent = transaction.prepare(agentOptions, session)
       await transaction.waitFor(options.setup?.(agent.ctx))
       transaction.assertActive()
       return transaction.publish('resume')

packages/core/agent-loop/src/tool-calls.ts

@@ -3,8 +3,9 @@
  * the assistant message's `tool-call` blocks; this module parses each call's
  * arguments once, classifies it via `ctx.tools.executionMode`, partitions the
  * calls into ordered groups (one exclusive call, or a run of consecutive
- * parallel-safe calls), and executes each group — a parallel group through a
- * rolling pool bounded by the agent's `maxParallelToolCalls`.
+ * parallel-safe calls), and runs every group through the same rolling pool
+ * bounded by the agent's `maxParallelToolCalls` — an exclusive group is a pool
+ * of one.
  *
  * The session log stays the source of truth and is reconstructable regardless
  * of dispatch timing: each STARTED call appends its own `tool/call` before its
@@ -95,16 +96,13 @@ export async function executeToolCalls(
   // separate ordered groups (no read/write race inside one assistant step).
   const groups = groupByMode(ctx, planned)
 
+  // Every group runs through the same rolling pool: an exclusive call is a
+  // singleton group (pool of one, a barrier), a parallel-safe run is one group
+  // bounded by the cap. `groupByMode` already classified each call, so the loop
+  // does not re-query `executionMode` here.
   const pendingContext: HookContext[] = []
   for (const group of groups) {
-    // Groups are never empty (groupByMode only pushes non-empty runs/singletons).
-    // eslint-disable-next-line @typescript-eslint/no-non-null-assertion -- non-empty group
-    const first = group[0]!
-    if (group.length === 1 && ctx.tools.executionMode(first.exec).kind === 'exclusive') {
-      await runExclusive(ctx, session, turn, step, first, signal, pendingContext)
-    } else {
-      await runParallelGroup(ctx, session, turn, step, group, signal, maxParallel, pendingContext)
-    }
+    await runGroup(ctx, session, turn, step, group, signal, maxParallel, pendingContext)
   }
   return pendingContext
 }
@@ -135,8 +133,8 @@ function parseArguments(raw: string): unknown {
 /**
  * Group planned calls into ordered runs: each exclusive call is a singleton
  * group; consecutive parallel-safe calls coalesce into one group. `executionMode`
- * is queried once per call here and again by the caller to pick the exclusive
- * fast-path — both reads are pure and cheap.
+ * is the sole classification point — the caller runs every group through the
+ * rolling pool without re-querying it. The read is pure and cheap.
  */
 function groupByMode(ctx: Context, planned: PlannedCall[]): PlannedCall[][] {
   const groups: PlannedCall[][] = []
@@ -167,46 +165,19 @@ function assertMaxParallelToolCalls(maxParallel: number): void {
 }
 
 /**
- * The exclusive single-call path keeps the public one-call pipeline sequential:
- * abort-check, `tool/call`, pre/dispatch/post via `ctx.tools.execute`,
- * `tool/result`, buffer context, post-await abort-check.
- */
-async function runExclusive(
-  ctx: Context,
-  session: Session,
-  turn: number,
-  step: number,
-  call: PlannedCall,
-  signal: AbortSignal,
-  pendingContext: HookContext[],
-): Promise<void> {
-  /* v8 ignore next -- signal.reason always set: cancel()/disposal provide a default */
-  if (signal.aborted) throw new Error(String(signal.reason ?? 'aborted'))
-  const callSeq = appendToolCall(session, turn, step, call.block)
-  const result = await ctx.tools.execute(call.exec)
-  appendToolResult(session, turn, step, call.block, result, callSeq)
-  if (result.additionalContext) pendingContext.push(result.additionalContext)
-  // signal CAN flip during the await above (abort() inside a tool); the analyzer
-  // can't see through the await boundary.
-  /* v8 ignore start -- signal.reason default unreachable: cancel()/disposal always set it */
-  // eslint-disable-next-line @typescript-eslint/no-unnecessary-condition
-  if (signal.aborted) throw new Error(String(signal.reason ?? 'aborted'))
-  /* v8 ignore stop */
-}
-
-/**
- * The rolling-pool path for a group of parallel-safe calls. Starts calls in
- * model order up to `maxParallel`, and whenever one settles starts the next
- * unstarted call until the group is exhausted. Settled dispatches land in
- * model-order slots; a commit cursor appends `tool/result` (and collects
- * `additionalContext`) only while the next slot is ready, so the log stays
- * model-ordered regardless of completion order.
+ * The rolling-pool path for one ordered group. A singleton exclusive group runs
+ * as a pool of one (a barrier); a parallel-safe run starts calls in model order
+ * up to `maxParallel`, and whenever one settles starts the next unstarted call
+ * until the group is exhausted. Settled dispatches land in model-order slots; a
+ * commit cursor appends `tool/result` (and collects `additionalContext`) only
+ * while the next slot is ready, so the log stays model-ordered regardless of
+ * completion order.
  *
  * Abort: an already-aborted signal starts nothing and throws before any
  * `tool/call`. An abort mid-group stops replenishment, awaits only the started
  * calls, commits their results in order, drops buffered context, and throws.
  */
-async function runParallelGroup(
+async function runGroup(
   ctx: Context,
   session: Session,
   turn: number,

packages/core/agent-loop/tests/tool-calls.spec.ts

@@ -274,6 +274,48 @@ describe('tool-call scheduler: rolling pool honors maxParallelToolCalls', () =>
     gated.release('2')
     await waitForIdle(ctx, agent)
   })
+
+  it('applies the factory-wide Config default to agents that set no per-agent cap', async () => {
+    const adapter = new MockAdapter([
+      multiCall([{ id: 'c1', name: 'p', args: { id: '1' } }, { id: 'c2', name: 'p', args: { id: '2' } }]),
+      textResponse('done'),
+    ])
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(SessionStore)
+    await ctx.plugin(SystemPrompt, { persona: '' })
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+    // Factory default of 1 (no per-agent cap set below) must serialize.
+    await ctx.plugin(AgentLoop, { agents: [], maxParallelToolCalls: 1 })
+    ctx.llm.registerAdapter(['mock'], adapter)
+    const gated = gatedParallelTool('p')
+    ctx.tools.register(gated.tool)
+    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock' })
+    expect(agent.options.maxParallelToolCalls).toBe(1)
+
+    agent.send([{ type: 'text', text: 'go' }])
+    await until(() => gated.started.length === 1)
+    await new Promise(r => setTimeout(r, 5))
+    expect(gated.started).toEqual(['1'])
+    gated.release('1')
+    await until(() => gated.started.length === 2)
+    gated.release('2')
+    await waitForIdle(ctx, agent)
+  })
+
+  it('lets a per-agent cap override the factory-wide Config default', async () => {
+    const ctx = new Context()
+    await ctx.plugin(LlmService)
+    await ctx.plugin(SessionStore)
+    await ctx.plugin(SystemPrompt, { persona: '' })
+    await ctx.plugin(ToolRegistry)
+    await ctx.plugin(AgentRegistry)
+    await ctx.plugin(AgentLoop, { agents: [], maxParallelToolCalls: 1 })
+    ctx.llm.registerAdapter(['mock'], new MockAdapter([]))
+    const agent = ctx.agentLoop.create(AgentId('a1'), { model: 'mock', maxParallelToolCalls: 4 })
+    expect(agent.options.maxParallelToolCalls).toBe(4)
+  })
 })
 
 describe('tool-call scheduler: ordered middleware and additionalContext', () => {