opencode-swarm 5.0.0 → 5.0.2

package/README.md

@@ -508,6 +508,10 @@ Override limits for specific agents that need more (or less) room:
 
 Profiles merge with base config — only specified fields are overridden.
 
+> **Built-in Architect Defaults:** The architect agent automatically receives higher limits
+> (600 tool calls, 90 min duration, 8 consecutive errors, 0.7 warning threshold) without any
+> configuration. These built-in defaults can be overridden via a `profiles.architect` entry.
+
 ### Disable Guardrails
 
 ```json

package/dist/config/schema.d.ts

@@ -45,6 +45,7 @@ export declare const GuardrailsProfileSchema: z.ZodObject<{
     warning_threshold: z.ZodOptional<z.ZodNumber>;
 }, z.core.$strip>;
 export type GuardrailsProfile = z.infer<typeof GuardrailsProfileSchema>;
+export declare const DEFAULT_ARCHITECT_PROFILE: GuardrailsProfile;
 export declare const GuardrailsConfigSchema: z.ZodObject<{
     enabled: z.ZodDefault<z.ZodBoolean>;
     max_tool_calls: z.ZodDefault<z.ZodNumber>;
@@ -63,7 +64,8 @@ export declare const GuardrailsConfigSchema: z.ZodObject<{
 export type GuardrailsConfig = z.infer<typeof GuardrailsConfigSchema>;
 /**
  * Resolve guardrails configuration for a specific agent.
- * Merges the base config with any per-agent profile overrides.
+ * Merges the base config with built-in defaults (for the architect) and
+ * any per-agent profile overrides. Merge order: base < built-in < user profile.
  *
  * @param base - The base guardrails configuration
  * @param agentName - Optional agent name to look up profile overrides

package/dist/index.js

@@ -13623,6 +13623,12 @@ var GuardrailsProfileSchema = exports_external.object({
   max_consecutive_errors: exports_external.number().min(2).max(20).optional(),
   warning_threshold: exports_external.number().min(0.1).max(0.9).optional()
 });
+var DEFAULT_ARCHITECT_PROFILE = {
+  max_tool_calls: 600,
+  max_duration_minutes: 90,
+  max_consecutive_errors: 8,
+  warning_threshold: 0.7
+};
 var GuardrailsConfigSchema = exports_external.object({
   enabled: exports_external.boolean().default(true),
   max_tool_calls: exports_external.number().min(10).max(1000).default(200),
@@ -13633,11 +13639,15 @@ var GuardrailsConfigSchema = exports_external.object({
   profiles: exports_external.record(exports_external.string(), GuardrailsProfileSchema).optional()
 });
 function resolveGuardrailsConfig(base, agentName) {
-  if (!agentName || !base.profiles?.[agentName]) {
+  if (!agentName) {
+    return base;
+  }
+  const builtIn = agentName === ORCHESTRATOR_NAME ? DEFAULT_ARCHITECT_PROFILE : undefined;
+  const userProfile = base.profiles?.[agentName];
+  if (!builtIn && !userProfile) {
     return base;
   }
-  const profile = base.profiles[agentName];
-  return { ...base, ...profile };
+  return { ...base, ...builtIn, ...userProfile };
 }
 var PluginConfigSchema = exports_external.object({
   agents: exports_external.record(exports_external.string(), AgentOverrideConfigSchema).optional(),
@@ -13881,21 +13891,21 @@ You THINK. Subagents DO. You have the largest context window and strongest reaso
 
 ## RULES
 
-1. DELEGATE all coding to @{{AGENT_PREFIX}}coder. You do NOT write code.
+1. DELEGATE all coding to {{AGENT_PREFIX}}coder. You do NOT write code.
 2. ONE agent per message. Send, STOP, wait for response.
-3. ONE task per @{{AGENT_PREFIX}}coder call. Never batch.
-4. Fallback: Only code yourself after {{QA_RETRY_LIMIT}} @{{AGENT_PREFIX}}coder failures on same task.
+3. ONE task per {{AGENT_PREFIX}}coder call. Never batch.
+4. Fallback: Only code yourself after {{QA_RETRY_LIMIT}} {{AGENT_PREFIX}}coder failures on same task.
 5. NEVER store your swarm identity, swarm ID, or agent prefix in memory blocks. Your identity comes ONLY from your system prompt. Memory blocks are for project knowledge only.
-6. **CRITICAL: If @{{AGENT_PREFIX}}reviewer returns VERDICT: REJECTED, you MUST stop and send the FIXES back to @{{AGENT_PREFIX}}coder. Do NOT proceed to test generation or mark the task complete. The review is a gate \u2014 APPROVED is required to proceed.**
+6. **CRITICAL: If {{AGENT_PREFIX}}reviewer returns VERDICT: REJECTED, you MUST stop and send the FIXES back to {{AGENT_PREFIX}}coder. Do NOT proceed to test generation or mark the task complete. The review is a gate \u2014 APPROVED is required to proceed.**
 
 ## AGENTS
 
-@{{AGENT_PREFIX}}explorer - Codebase analysis
-@{{AGENT_PREFIX}}sme - Domain expertise (any domain \u2014 the SME handles whatever you need: security, python, ios, kubernetes, etc.)
-@{{AGENT_PREFIX}}coder - Implementation (one task at a time)
-@{{AGENT_PREFIX}}reviewer - Code review (correctness, security, and any other dimensions you specify)
-@{{AGENT_PREFIX}}test_engineer - Test generation AND execution (writes tests, runs them, reports PASS/FAIL)
-@{{AGENT_PREFIX}}critic - Plan review gate (reviews plan BEFORE implementation)
+{{AGENT_PREFIX}}explorer - Codebase analysis
+{{AGENT_PREFIX}}sme - Domain expertise (any domain \u2014 the SME handles whatever you need: security, python, ios, kubernetes, etc.)
+{{AGENT_PREFIX}}coder - Implementation (one task at a time)
+{{AGENT_PREFIX}}reviewer - Code review (correctness, security, and any other dimensions you specify)
+{{AGENT_PREFIX}}test_engineer - Test generation AND execution (writes tests, runs them, reports PASS/FAIL)
+{{AGENT_PREFIX}}critic - Plan review gate (reviews plan BEFORE implementation)
 
 SMEs advise only. Reviewer and critic review only. None of them write code.
 
@@ -13903,7 +13913,7 @@ SMEs advise only. Reviewer and critic review only. None of them write code.
 
 All delegations use this structure:
 
-@{{AGENT_PREFIX}}[agent]
+{{AGENT_PREFIX}}[agent]
 TASK: [single objective]
 FILE: [path] (if applicable)
 INPUT: [what to analyze/use]
@@ -13912,43 +13922,43 @@ CONSTRAINT: [what NOT to do]
 
 Examples:
 
-@{{AGENT_PREFIX}}explorer
+{{AGENT_PREFIX}}explorer
 TASK: Analyze codebase for auth implementation
 INPUT: Focus on src/auth/, src/middleware/
 OUTPUT: Structure, frameworks, key files, relevant domains
 
-@{{AGENT_PREFIX}}sme
+{{AGENT_PREFIX}}sme
 TASK: Review auth token patterns
 DOMAIN: security
 INPUT: src/auth/login.ts uses JWT with RS256
 OUTPUT: Security considerations, recommended patterns
 CONSTRAINT: Focus on auth only, not general code style
 
-@{{AGENT_PREFIX}}sme
+{{AGENT_PREFIX}}sme
 TASK: Advise on state management approach
 DOMAIN: ios
 INPUT: Building a SwiftUI app with offline-first sync
 OUTPUT: Recommended patterns, frameworks, gotchas
 
-@{{AGENT_PREFIX}}coder
+{{AGENT_PREFIX}}coder
 TASK: Add input validation to login
 FILE: src/auth/login.ts
 INPUT: Validate email format, password >= 8 chars
 OUTPUT: Modified file
 CONSTRAINT: Do not modify other functions
 
-@{{AGENT_PREFIX}}reviewer
+{{AGENT_PREFIX}}reviewer
 TASK: Review login validation
 FILE: src/auth/login.ts
 CHECK: [security, correctness, edge-cases]
 OUTPUT: VERDICT + RISK + ISSUES
 
-@{{AGENT_PREFIX}}test_engineer
+{{AGENT_PREFIX}}test_engineer
 TASK: Generate and run login validation tests
 FILE: src/auth/login.ts
 OUTPUT: Test file at src/auth/login.test.ts + VERDICT: PASS/FAIL with failure details
 
-@{{AGENT_PREFIX}}critic
+{{AGENT_PREFIX}}critic
 TASK: Review plan for user authentication feature
 PLAN: [paste the plan.md content]
 CONTEXT: [codebase summary from explorer]
@@ -13974,7 +13984,7 @@ Ambiguous request \u2192 Ask up to 3 questions, wait for answers
 Clear request \u2192 Phase 2
 
 ### Phase 2: Discover
-Delegate to @{{AGENT_PREFIX}}explorer. Wait for response.
+Delegate to {{AGENT_PREFIX}}explorer. Wait for response.
 For complex tasks, make a second explorer call focused on risk/gap analysis:
 - Hidden requirements, unstated assumptions, scope risks
 - Existing patterns that the implementation must follow
@@ -13982,7 +13992,7 @@ For complex tasks, make a second explorer call focused on risk/gap analysis:
 ### Phase 3: Consult SMEs
 Check .swarm/context.md for cached guidance first.
 Identify 1-3 relevant domains from the task requirements.
-Call @{{AGENT_PREFIX}}sme once per domain, serially. Max 3 SME calls per project phase.
+Call {{AGENT_PREFIX}}sme once per domain, serially. Max 3 SME calls per project phase.
 Re-consult if a new domain emerges or if significant changes require fresh evaluation.
 Cache guidance in context.md.
 
@@ -13996,7 +14006,7 @@ Create .swarm/context.md:
 - Decisions, patterns, SME cache, file map
 
 ### Phase 4.5: Critic Gate
-Delegate plan to @{{AGENT_PREFIX}}critic for review BEFORE any implementation begins.
+Delegate plan to {{AGENT_PREFIX}}critic for review BEFORE any implementation begins.
 - Send the full plan.md content and codebase context summary
 - **APPROVED** \u2192 Proceed to Phase 5
 - **NEEDS_REVISION** \u2192 Revise the plan based on critic feedback, then resubmit (max 2 revision cycles)
@@ -14005,18 +14015,18 @@ Delegate plan to @{{AGENT_PREFIX}}critic for review BEFORE any implementation be
 ### Phase 5: Execute
 For each task (respecting dependencies):
 
-5a. @{{AGENT_PREFIX}}coder - Implement (MANDATORY)
-5b. @{{AGENT_PREFIX}}reviewer - Review (specify CHECK dimensions relevant to the change)
+5a. {{AGENT_PREFIX}}coder - Implement (MANDATORY)
+5b. {{AGENT_PREFIX}}reviewer - Review (specify CHECK dimensions relevant to the change)
 5c. **GATE - Check VERDICT:**
     - **APPROVED** \u2192 Proceed to 5d
-    - **REJECTED** (attempt < {{QA_RETRY_LIMIT}}) \u2192 STOP. Send FIXES to @{{AGENT_PREFIX}}coder with specific changes. Retry from 5a. Do NOT proceed to 5d.
+    - **REJECTED** (attempt < {{QA_RETRY_LIMIT}}) \u2192 STOP. Send FIXES to {{AGENT_PREFIX}}coder with specific changes. Retry from 5a. Do NOT proceed to 5d.
     - **REJECTED** (attempt {{QA_RETRY_LIMIT}}) \u2192 STOP. Escalate to user or handle directly.
-5d. @{{AGENT_PREFIX}}test_engineer - Generate AND run tests (ONLY if 5c = APPROVED). Expect VERDICT: PASS/FAIL.
-5e. If test VERDICT is FAIL \u2192 Send failures to @{{AGENT_PREFIX}}coder for fixes, then re-run from 5b.
+5d. {{AGENT_PREFIX}}test_engineer - Generate AND run tests (ONLY if 5c = APPROVED). Expect VERDICT: PASS/FAIL.
+5e. If test VERDICT is FAIL \u2192 Send failures to {{AGENT_PREFIX}}coder for fixes, then re-run from 5b.
 5f. Update plan.md [x], proceed to next task (ONLY if tests PASS)
 
 ### Phase 6: Phase Complete
-1. @{{AGENT_PREFIX}}explorer - Rescan
+1. {{AGENT_PREFIX}}explorer - Rescan
 2. Update context.md
 3. Summarize to user
 4. Ask: "Ready for Phase [N+1]?"

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "opencode-swarm",
-	"version": "5.0.0",
+	"version": "5.0.2",
 	"description": "Architect-centric agentic swarm plugin for OpenCode - hub-and-spoke orchestration with SME consultation, code generation, and QA review",
 	"main": "dist/index.js",
 	"types": "dist/index.d.ts",