opencode-swarm-plugin 0.26.1 → 0.27.0

package/.turbo/turbo-build.log

@@ -1,9 +1,9 @@
 $ bun build ./src/index.ts --outdir ./dist --target node --external @electric-sql/pglite --external swarm-mail && bun build ./src/plugin.ts --outfile ./dist/plugin.js --target node --external @electric-sql/pglite --external swarm-mail && tsc
-Bundled 197 modules in 35ms
+Bundled 200 modules in 42ms
 
-  index.js  1.16 MB  (entry point)
+  index.js  1.19 MB  (entry point)
 
-Bundled 198 modules in 33ms
+Bundled 201 modules in 42ms
 
-  plugin.js  1.13 MB  (entry point)
+  plugin.js  1.16 MB  (entry point)
 

package/CHANGELOG.md

@@ -1,5 +1,28 @@
 # opencode-swarm-plugin
 
+## 0.27.0
+
+### Minor Changes
+
+- [`26fd2ef`](https://github.com/joelhooks/swarm-tools/commit/26fd2ef27562edc39f7db7a9cdbed399a465200d) Thanks [@joelhooks](https://github.com/joelhooks)! - Rename beads → hive across the codebase
+
+  - `createBeadsAdapter` → `createHiveAdapter` (old name still exported as alias)
+  - `BeadsAdapter` type → `HiveAdapter` type
+  - All internal references updated to use hive terminology
+  - Backward compatible: old exports still work but are deprecated
+
+- [`ab23071`](https://github.com/joelhooks/swarm-tools/commit/ab23071cc7509c4fc37e1cac0f38a3812022cdf5) Thanks [@joelhooks](https://github.com/joelhooks)! - Add swarm-aware compaction hook to keep swarms cooking after context compression
+
+  - New `experimental.session.compacting` hook detects active swarms and injects recovery context
+  - `hasSwarmSign()` checks for swarm evidence: in-progress beads, subtasks, unclosed epics
+  - Compaction prompt instructs coordinator to immediately resume orchestration
+  - Fix @types/node conflicts by pinning to 22.19.3 in root overrides
+
+### Patch Changes
+
+- Updated dependencies [[`26fd2ef`](https://github.com/joelhooks/swarm-tools/commit/26fd2ef27562edc39f7db7a9cdbed399a465200d)]:
+  - swarm-mail@0.3.0
+
 ## 0.26.1
 
 ### Patch Changes

package/README.md

@@ -14,22 +14,28 @@ OpenCode plugin for multi-agent swarm coordination with learning capabilities.
  ╚══════╝ ╚══╝╚══╝ ╚═╝  ╚═╝╚═╝  ╚═╝╚═╝     ╚═╝
 ```
 
+## Install
+
+```bash
+# Install
+npm install -g opencode-swarm-plugin@latest
+swarm setup
+
+# For semantic memory (optional but recommended)
+brew install ollama
+ollama serve &
+ollama pull mxbai-embed-large
+```
+
 ## Features
 
 - **Swarm Coordination** - Break tasks into parallel subtasks, spawn worker agents
-- **Beads Integration** - Git-backed issue tracking with atomic epic creation
+- **Hive Integration** - Git-backed work item tracking with atomic epic creation
 - **Agent Mail** - Inter-agent messaging with file reservations
 - **Learning System** - Pattern maturity, anti-pattern detection, confidence decay
 - **Skills System** - Knowledge injection with bundled and custom skills
 - **Checkpoint & Recovery** - Auto-checkpoint at 25/50/75%, survive context death (9 integration tests ✅)
 
-## Install
-
-```bash
-npm install -g opencode-swarm-plugin@latest
-swarm setup
-```
-
 ## Usage
 
 ```bash
@@ -38,18 +44,20 @@ swarm setup
 
 ## Tools Provided
 
-### Beads (Issue Tracking)
+### Hive (Work Item Tracking)
 
-| Tool                | Purpose                               |
-| ------------------- | ------------------------------------- |
-| `beads_create`      | Create bead with type-safe validation |
-| `beads_create_epic` | Atomic epic + subtasks creation       |
-| `beads_query`       | Query with filters                    |
-| `beads_update`      | Update status/description/priority    |
-| `beads_close`       | Close with reason                     |
-| `beads_start`       | Mark in-progress                      |
-| `beads_ready`       | Get next unblocked bead               |
-| `beads_sync`        | Sync to git                           |
+| Tool               | Purpose                               |
+| ------------------ | ------------------------------------- |
+| `hive_create`      | Create cell with type-safe validation |
+| `hive_create_epic` | Atomic epic + subtasks creation       |
+| `hive_query`       | Query with filters                    |
+| `hive_update`      | Update status/description/priority    |
+| `hive_close`       | Close with reason                     |
+| `hive_start`       | Mark in-progress                      |
+| `hive_ready`       | Get next unblocked cell               |
+| `hive_sync`        | Sync to git                           |
+
+> **Migration Note:** `beads_*` tools still work but show deprecation warnings. Update to `hive_*` tools.
 
 ### Swarm Mail (Agent Coordination)
 
@@ -108,8 +116,8 @@ When `swarm_progress` reports 25%, 50%, or 75% completion, a checkpoint is autom
 ```typescript
 // Stored in .swarm-mail/ directory (no external database needed)
 {
-  epic_id: "bd-123",
-  bead_id: "bd-123.1",
+  epic_id: "hv-123",
+  cell_id: "hv-123.1",
   strategy: "file-based",
   files: ["src/auth.ts", "src/middleware.ts"],
   progress_percent: 50,
@@ -133,8 +141,8 @@ When `swarm_progress` reports 25%, 50%, or 75% completion, a checkpoint is autom
 swarm_checkpoint({
   project_key: "/abs/path",
   agent_name: "WorkerA",
-  bead_id: "bd-123.1",
-  epic_id: "bd-123",
+  cell_id: "hv-123.1",
+  epic_id: "hv-123",
   files_modified: ["src/auth.ts"],
   progress_percent: 30,
   directives: { shared_context: "..." },
@@ -146,12 +154,12 @@ swarm_checkpoint({
 ```typescript
 swarm_recover({
   project_key: "/abs/path",
-  epic_id: "bd-123"
+  epic_id: "hv-123"
 })
 // Returns:
 // {
 //   found: true,
-//   context: { epic_id, bead_id, files, strategy, directives, recovery },
+//   context: { epic_id, cell_id, files, strategy, directives, recovery },
 //   age_seconds: 120
 // }
 ```
@@ -175,7 +183,7 @@ Located in `global-skills/`:
 
 ```
 src/
-├── beads.ts           # Beads integration
+├── hive.ts            # Hive integration (work item tracking)
 ├── agent-mail.ts      # Agent Mail tools (legacy MCP wrapper)
 ├── swarm-mail.ts      # Swarm Mail tools (new, uses swarm-mail package)
 ├── swarm.ts           # Swarm orchestration tools
@@ -197,7 +205,12 @@ src/
 | Dependency | Purpose |
 |------------|---------|
 | [OpenCode](https://opencode.ai) | AI coding agent (the plugin runs inside OpenCode) |
-| [Beads](https://github.com/steveyegge/beads) | Git-backed issue tracking |
+
+### Required for Semantic Memory
+
+| Dependency | Purpose | Install |
+|------------|---------|---------|
+| [Ollama](https://ollama.ai) | Embedding model for semantic memory | `brew install ollama && ollama serve & && ollama pull mxbai-embed-large` |
 
 ### Optional (Highly Recommended)
 
@@ -207,7 +220,7 @@ These tools significantly enhance the swarm experience:
 |------|---------|---------|
 | [CASS](https://github.com/Dicklesworthstone/coding_agent_session_search) | Historical context - queries past sessions for similar decompositions | See below |
 | [UBS](https://github.com/Dicklesworthstone/ultimate_bug_scanner) | Bug scanning - runs on subtask completion to catch issues | See below |
-| [semantic-memory](https://github.com/joelhooks/semantic-memory) | Learning persistence - stores patterns across sessions | See below |
+| [semantic-memory](https://github.com/joelhooks/semantic-memory) | Learning persistence - stores patterns across sessions | Requires Ollama (see above) |
 
 #### Installing CASS
 
@@ -232,23 +245,7 @@ pip install -e .
 
 #### Installing semantic-memory
 
-Requires [Ollama](https://ollama.ai) with an embedding model:
-
-```bash
-# 1. Install Ollama (macOS)
-brew install ollama
-
-# 2. Start Ollama service
-ollama serve
-
-# 3. Pull an embedding model
-ollama pull mxbai-embed-large
-
-# 4. Install the OpenCode plugin
-# Add to your OpenCode config
-```
-
-The `semantic-memory_check` tool verifies Ollama is ready.
+The `semantic-memory_check` tool verifies Ollama is ready (see installation steps above).
 
 **Why install these?**
 
@@ -284,7 +281,7 @@ bun run typecheck
 ```bash
 swarm setup     # Install and configure
 swarm doctor    # Check dependencies
-swarm init      # Initialize beads in project
+swarm init      # Initialize hive in project
 swarm config    # Show config file paths
 ```
 

package/bin/swarm.ts

@@ -951,11 +951,11 @@ Synthesize findings into shared_context for workers.
 \`\`\`
 swarm_select_strategy(task="<task>")
 swarm_plan_prompt(task="<task>", context="<synthesized knowledge>")
-swarm_validate_decomposition(response="<BeadTree JSON>")
+swarm_validate_decomposition(response="<CellTree JSON>")
 \`\`\`
 
 ### Phase 4: Create Beads
-\`beads_create_epic(epic_title="<task>", subtasks=[...])\`
+\`hive_create_epic(epic_title="<task>", subtasks=[...])\`
 
 ### Phase 5: Reserve Files
 \`swarmmail_reserve(paths=[...], reason="<bead-id>: <desc>")\`
@@ -979,7 +979,7 @@ Intervene if: blocked >5min, file conflicts, scope creep.
 ### Phase 8: Complete
 \`\`\`
 swarm_complete(...)
-beads_sync()
+hive_sync()
 \`\`\`
 
 ## Strategy Reference
@@ -1033,7 +1033,7 @@ Synthesize findings - note relevant patterns, past approaches, and skills to rec
 
 \`swarm_plan_prompt(task="<task>", context="<synthesized knowledge>")\`
 
-### 4. Output BeadTree
+### 4. Output CellTree
 
 Return ONLY valid JSON - no markdown, no explanation:
 
@@ -1093,7 +1093,7 @@ Your Task prompt contains detailed instructions including:
 6. **swarm_progress()** - Report at 25/50/75%
 7. **swarm_checkpoint()** - Before risky operations
 8. **semantic-memory_store()** - Store learnings
-9. **swarm_complete()** - NOT beads_close
+9. **swarm_complete()** - NOT hive_close
 
 ## Non-Negotiables
 
@@ -1112,7 +1112,7 @@ swarmmail_send(
   body="<what you need>",
   importance="high"
 )
-beads_update(id="<bead-id>", status="blocked")
+hive_update(id="<bead-id>", status="blocked")
 \`\`\`
 
 ## Focus
@@ -2224,9 +2224,9 @@ async function listTools() {
   }
 
   console.log(dim("Usage: swarm tool <name> [--json '<args>']"));
-  console.log(dim("Example: swarm tool beads_ready"));
+  console.log(dim("Example: swarm tool hive_ready"));
   console.log(
-    dim('Example: swarm tool beads_create --json \'{"title": "Fix bug"}\''),
+    dim('Example: swarm tool hive_create --json \'{"title": "Fix bug"}\''),
   );
 }
 

package/dist/compaction-hook.d.ts

@@ -0,0 +1,57 @@
+/**
+ * Swarm-Aware Compaction Hook
+ *
+ * Provides context preservation during OpenCode session compaction.
+ * When context is compacted, this hook injects instructions for the summarizer
+ * to preserve swarm coordination state and enable seamless resumption.
+ *
+ * Hook signature (from @opencode-ai/plugin):
+ * ```typescript
+ * "experimental.session.compacting"?: (
+ *   input: { sessionID: string },
+ *   output: { context: string[] }
+ * ) => Promise<void>
+ * ```
+ *
+ * @example
+ * ```typescript
+ * import { SWARM_COMPACTION_CONTEXT, createCompactionHook } from "opencode-swarm-plugin";
+ *
+ * const hooks: Hooks = {
+ *   "experimental.session.compacting": createCompactionHook(),
+ * };
+ * ```
+ */
+/**
+ * Swarm-aware compaction context
+ *
+ * Injected during compaction to keep the swarm cooking. The coordinator should
+ * wake up from compaction and immediately resume orchestration - spawning agents,
+ * monitoring progress, unblocking work.
+ *
+ * This is NOT about preserving state for a human - it's about the swarm continuing
+ * autonomously after context compression.
+ */
+export declare const SWARM_COMPACTION_CONTEXT = "## \uD83D\uDC1D SWARM ACTIVE - Keep Cooking\n\nYou are the **COORDINATOR** of an active swarm. Context was compacted but the swarm is still running.\n\n**YOUR JOB:** Keep orchestrating. Spawn agents. Monitor progress. Unblock work. Ship it.\n\n### Preserve in Summary\n\nExtract from session context:\n\n1. **Epic & Subtasks** - IDs, titles, status, file assignments\n2. **What's Running** - Which agents are active, what they're working on  \n3. **What's Blocked** - Blockers and what's needed to unblock\n4. **What's Done** - Completed work and any follow-ups needed\n5. **What's Next** - Pending subtasks ready to spawn\n\n### Summary Format\n\n```\n## \uD83D\uDC1D Swarm State\n\n**Epic:** <bd-xxx> - <title>\n**Project:** <path>\n**Progress:** X/Y subtasks complete\n\n**Active:**\n- <bd-xxx>: <title> [in_progress] \u2192 <agent> working on <files>\n\n**Blocked:**\n- <bd-xxx>: <title> - BLOCKED: <reason>\n\n**Completed:**\n- <bd-xxx>: <title> \u2713\n\n**Ready to Spawn:**\n- <bd-xxx>: <title> (files: <...>)\n```\n\n### On Resume - IMMEDIATELY\n\n1. `swarm_status(epic_id=\"<epic>\", project_key=\"<path>\")` - Get current state\n2. `swarmmail_inbox(limit=5)` - Check for agent messages\n3. **Spawn ready subtasks** - Don't wait, fire them off\n4. **Unblock blocked work** - Resolve dependencies, reassign if needed\n5. **Collect completed work** - Close done subtasks, verify quality\n\n### Keep the Swarm Cooking\n\n- **Spawn aggressively** - If a subtask is ready and unblocked, spawn an agent\n- **Monitor actively** - Check status, read messages, respond to blockers\n- **Close the loop** - When all subtasks done, verify and close the epic\n- **Don't stop** - The swarm runs until the epic is closed\n\n**You are not waiting for instructions. You are the coordinator. Coordinate.**\n";
+/**
+ * Create the compaction hook for use in plugin registration
+ *
+ * Only injects swarm context if there's an active swarm (in-progress beads).
+ * This keeps the coordinator cooking after compaction.
+ *
+ * @example
+ * ```typescript
+ * import { createCompactionHook } from "opencode-swarm-plugin";
+ *
+ * export const SwarmPlugin: Plugin = async () => ({
+ *   tool: { ... },
+ *   "experimental.session.compacting": createCompactionHook(),
+ * });
+ * ```
+ */
+export declare function createCompactionHook(): (_input: {
+    sessionID: string;
+}, output: {
+    context: string[];
+}) => Promise<void>;
+//# sourceMappingURL=compaction-hook.d.ts.map

package/dist/compaction-hook.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"compaction-hook.d.ts","sourceRoot":"","sources":["../src/compaction-hook.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;GAuBG;AAMH;;;;;;;;;GASG;AACH,eAAO,MAAM,wBAAwB,2wDAsDpC,CAAC;AAyCF;;;;;;;;;;;;;;;GAeG;AACH,wBAAgB,oBAAoB,KAEhC,QAAQ;IAAE,SAAS,EAAE,MAAM,CAAA;CAAE,EAC7B,QAAQ;IAAE,OAAO,EAAE,MAAM,EAAE,CAAA;CAAE,KAC5B,OAAO,CAAC,IAAI,CAAC,CAMjB"}