Merge pull request #553 from bradygaster/dev

Release v0.9.0

Author: bradygaster

.changeset/add-rework-rate-metric.md

@@ -0,0 +1,5 @@
+---
+"@bradygaster/squad-sdk": minor
+---
+
+Add rework rate OTEL metrics (#265) — tracks PR revision patterns as the emerging 5th DORA metric

.changeset/add-session-recovery-skill.md

@@ -0,0 +1,4 @@
+---
+---
+
+Add session-recovery skill for detecting and resuming interrupted Copilot CLI sessions

.changeset/add-skill-script-loader.md

@@ -0,0 +1,5 @@
+---
+"@bradygaster/squad-sdk": minor
+---
+
+Add SkillScriptLoader, handler types, and applySkillHandlers for executable skill handler scripts

.changeset/ado-configurable-items.md

@@ -0,0 +1,14 @@
+---
+"@bradygaster/squad-sdk": minor
+---
+
+ADO configurable work item types, area paths, and iteration support (#240)
+
+- Add `getAvailableWorkItemTypes(org, project)` for process template introspection
+- Add `validateWorkItemType(org, project, typeName)` for pre-creation validation
+- Add `WorkItemTypeInfo` interface for type metadata
+- Enhance `AzureDevOpsAdapter` with instance-level `getAvailableWorkItemTypes()` and `validateWorkItemType()` methods
+- Support optional `validateType` flag in `createWorkItem()` to check type before creation
+- Enhance `squad init` to introspect available work item types and store `_availableTypes` hint in `.squad/config.json`
+- Support `adoConfig` in `InitOptions` for explicit work item type, area path, and iteration path during init
+- Graceful fallback to default types (User Story, Bug, Task) when az CLI is unavailable

.changeset/config.json

@@ -5,7 +5,7 @@
   "fixed": [],
   "linked": [],
   "access": "public",
-  "baseBranch": "main",
+  "baseBranch": "dev",
   "updateInternalDependencies": "patch",
   "ignore": []
 }

.changeset/cooperative-rate-limiting.md

@@ -0,0 +1,11 @@
+---
+"@bradygaster/squad-sdk": minor
+---
+
+feat: Cooperative rate limiting with predictive circuit breaker
+
+Added cooperative rate limiting patterns for multi-agent deployments:
+- Traffic Light, Predictive Circuit Breaker, Priority Retry Windows
+- Cooperative Token Pool for shared quota management
+
+Closes #515

.changeset/cross-squad-orchestration.md

@@ -0,0 +1,6 @@
+---
+"@bradygaster/squad-sdk": minor
+"@bradygaster/squad-cli": minor
+---
+
+Cross-squad orchestration — discovery, delegation, and manifest (#316). Adds manifest schema, `squad discover` / `squad delegate` commands, and runtime module for coordinating work across multiple Squad instances.

.changeset/fix-readme-migration-guide-link.md

@@ -0,0 +1,5 @@
+---
+"@bradygaster/squad-cli": patch
+---
+
+docs: point README migration guide links to the published docs site

.changeset/keda-scaler-docs.md

@@ -0,0 +1,10 @@
+---
+"@bradygaster/squad-cli": patch
+---
+
+docs: Add KEDA external scaler template for GitHub issue-driven autoscaling
+
+New template documenting how to use keda-copilot-scaler for scaling
+agent pods to zero when idle, with rate-limit-aware polling.
+
+Closes #516

.changeset/machine-capability-routing.md

@@ -0,0 +1,15 @@
+---
+"@bradygaster/squad-sdk": minor
+"@bradygaster/squad-cli": minor
+---
+
+feat: Machine capability discovery and `needs:*` label-based issue routing
+
+Added capability filtering to Ralph's watch command. Issues with `needs:*` labels
+(e.g., `needs:gpu`, `needs:browser`) are only processed by Ralph instances whose
+machine has those capabilities declared in `machine-capabilities.json`.
+
+This enables multi-machine Squad deployments where different machines handle
+different types of work based on their available tooling.
+
+Closes #514