opencode-swarm 6.84.0 → 6.84.1

package/README.md

@@ -17,9 +17,11 @@
 OpenCode Swarm is a plugin for [OpenCode](https://opencode.ai) that turns a single AI coding session into an **architect-led team of 11 specialized agents**. One agent writes the code. A different agent reviews it. Another writes and runs tests. Another checks security. **Nothing ships until every required gate passes.**
 
 ```bash
-npm install -g opencode-swarm
+bunx opencode-swarm install
 ```
 
+> This single command installs the package, registers it as an OpenCode plugin, disables conflicting default agents, and creates a ready-to-edit config at `~/.config/opencode/opencode-swarm.json`. Requires [Bun](https://bun.sh) (`bun --version` to check). If you must use npm: `npm install -g opencode-swarm && opencode-swarm install`.
+
 ### Why Swarm?
 
 Most AI coding tools let one model write code and ask that same model whether the code is good. That misses too much. Swarm separates planning, implementation, review, testing, and documentation into specialized internal roles — and enforces gated execution so agents never mutate the codebase in parallel.

package/dist/cli/index.js

@@ -19268,21 +19268,25 @@ var DEFAULT_AGENT_PROFILES = {
   coder: {
     max_tool_calls: 400,
     max_duration_minutes: 45,
+    max_consecutive_errors: 8,
     warning_threshold: 0.85
   },
   test_engineer: {
     max_tool_calls: 400,
     max_duration_minutes: 45,
+    max_consecutive_errors: 8,
     warning_threshold: 0.85
   },
   explorer: {
     max_tool_calls: 150,
     max_duration_minutes: 20,
+    max_consecutive_errors: 8,
     warning_threshold: 0.75
   },
   reviewer: {
     max_tool_calls: 200,
     max_duration_minutes: 30,
+    max_consecutive_errors: 8,
     warning_threshold: 0.65
   },
   critic: {
@@ -44320,7 +44324,7 @@ async function handleResetSessionCommand(directory, _args) {
     "",
     "Session state cleared. Plan, evidence, and knowledge preserved.",
     "",
-    "**Next step:** Start a new OpenCode session. The plugin will initialize fresh session state on startup."
+    "**All circuit breakers and revision limits have been cleared.** You can continue in this session \u2014 fresh state will be initialized automatically on the next tool call."
   ].join(`
 `);
 }

package/dist/index.js

@@ -15016,21 +15016,25 @@ var init_schema = __esm(() => {
     coder: {
       max_tool_calls: 400,
       max_duration_minutes: 45,
+      max_consecutive_errors: 8,
       warning_threshold: 0.85
     },
     test_engineer: {
       max_tool_calls: 400,
       max_duration_minutes: 45,
+      max_consecutive_errors: 8,
       warning_threshold: 0.85
     },
     explorer: {
       max_tool_calls: 150,
       max_duration_minutes: 20,
+      max_consecutive_errors: 8,
       warning_threshold: 0.75
     },
     reviewer: {
       max_tool_calls: 200,
       max_duration_minutes: 30,
+      max_consecutive_errors: 8,
       warning_threshold: 0.65
     },
     critic: {
@@ -23380,7 +23384,7 @@ function createGuardrailsHooks(directory, directoryOrConfig, config2, authorityC
     if (window2.consecutiveErrors >= agentConfig.max_consecutive_errors) {
       window2.hardLimitHit = true;
       telemetry.hardLimitHit(sessionID, window2.agentName, "consecutive_errors", window2.consecutiveErrors);
-      throw new Error(`\uD83D\uDED1 LIMIT REACHED: ${window2.consecutiveErrors} consecutive tool errors detected. Return your progress summary with details of what went wrong.`);
+      throw new Error(`\uD83D\uDED1 LIMIT REACHED: ${window2.consecutiveErrors} consecutive tool errors detected. Return your progress summary with details of what went wrong. Run /swarm reset-session to clear the circuit breaker without restarting your session.`);
     }
     const idleMinutes = (Date.now() - window2.lastSuccessTimeMs) / 60000;
     if (idleMinutes >= agentConfig.idle_timeout_minutes) {
@@ -23921,31 +23925,32 @@ function createGuardrailsHooks(directory, directoryOrConfig, config2, authorityC
         return;
       const hasError = output.output === null || output.output === undefined;
       if (hasError) {
-        window2.consecutiveErrors++;
-        if (session) {
-          const outputStr = typeof output.output === "string" ? output.output : "";
-          const errorContent = output.error ?? outputStr;
-          if (typeof errorContent === "string" && TRANSIENT_MODEL_ERROR_PATTERN.test(errorContent) && !session.modelFallbackExhausted) {
-            session.model_fallback_index++;
-            const baseAgentName = session.agentName ? session.agentName.replace(/^[^_]+[_]/, "") : "";
-            const swarmAgents = getSwarmAgents();
-            const fallbackModels = swarmAgents?.[baseAgentName]?.fallback_models;
-            session.modelFallbackExhausted = !fallbackModels || session.model_fallback_index > fallbackModels.length;
-            const fallbackModel = resolveFallbackModel(baseAgentName, session.model_fallback_index, swarmAgents);
-            const primaryModel = swarmAgents?.[baseAgentName]?.model ?? "default";
-            if (fallbackModel) {
-              if (swarmAgents?.[baseAgentName]) {
-                swarmAgents[baseAgentName].model = fallbackModel;
-              }
-              session.pendingAdvisoryMessages ??= [];
-              session.pendingAdvisoryMessages.push(`MODEL FALLBACK: Applied fallback model "${fallbackModel}" (attempt ${session.model_fallback_index}). ` + `Using /swarm handoff to reset to primary model.`);
-            } else {
-              session.pendingAdvisoryMessages ??= [];
-              session.pendingAdvisoryMessages.push(`MODEL FALLBACK: Transient model error detected (attempt ${session.model_fallback_index}). ` + `No fallback models configured for this agent. Add "fallback_models": ["model-a", "model-b"] ` + `to the agent's config in opencode-swarm.json.`);
+        const outputStr = typeof output.output === "string" ? output.output : "";
+        const errorContent = output.error ?? outputStr;
+        const isTransient = !!session && !session.modelFallbackExhausted && typeof errorContent === "string" && TRANSIENT_MODEL_ERROR_PATTERN.test(errorContent);
+        if (!isTransient) {
+          window2.consecutiveErrors++;
+        }
+        if (session && isTransient) {
+          session.model_fallback_index++;
+          const baseAgentName = session.agentName ? session.agentName.replace(/^[^_]+[_]/, "") : "";
+          const swarmAgents = getSwarmAgents();
+          const fallbackModels = swarmAgents?.[baseAgentName]?.fallback_models;
+          session.modelFallbackExhausted = !fallbackModels || session.model_fallback_index > fallbackModels.length;
+          const fallbackModel = resolveFallbackModel(baseAgentName, session.model_fallback_index, swarmAgents);
+          const primaryModel = swarmAgents?.[baseAgentName]?.model ?? "default";
+          if (fallbackModel) {
+            if (swarmAgents?.[baseAgentName]) {
+              swarmAgents[baseAgentName].model = fallbackModel;
             }
-            telemetry.modelFallback(input.sessionID, session.agentName, primaryModel, fallbackModel ?? "none", "transient_model_error");
-            swarmState.pendingEvents++;
+            session.pendingAdvisoryMessages ??= [];
+            session.pendingAdvisoryMessages.push(`MODEL FALLBACK: Applied fallback model "${fallbackModel}" (attempt ${session.model_fallback_index}). ` + `Using /swarm handoff to reset to primary model.`);
+          } else {
+            session.pendingAdvisoryMessages ??= [];
+            session.pendingAdvisoryMessages.push(`MODEL FALLBACK: Transient model error detected (attempt ${session.model_fallback_index}). ` + `No fallback models configured for this agent. Add "fallback_models": ["model-a", "model-b"] ` + `to the agent's config in opencode-swarm.json.`);
           }
+          telemetry.modelFallback(input.sessionID, session.agentName, primaryModel, fallbackModel ?? "none", "transient_model_error");
+          swarmState.pendingEvents++;
         }
       } else {
         window2.consecutiveErrors = 0;
@@ -53076,7 +53081,7 @@ async function handleResetSessionCommand(directory, _args) {
     "",
     "Session state cleared. Plan, evidence, and knowledge preserved.",
     "",
-    "**Next step:** Start a new OpenCode session. The plugin will initialize fresh session state on startup."
+    "**All circuit breakers and revision limits have been cleared.** You can continue in this session \u2014 fresh state will be initialized automatically on the next tool call."
   ].join(`
 `);
 }

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "opencode-swarm",
-	"version": "6.84.0",
+	"version": "6.84.1",
 	"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",