@techwavedev/agi-agent-kit 1.3.5 → 1.4.2

package/templates/skills/extended/documentation/documentation-generation-doc-generate/SKILL.md

@@ -55,15 +55,18 @@ $ARGUMENTS
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -71,14 +74,16 @@ python3 execution/memory_manager.py store \
   --tags documentation-generation-doc-generate <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/docusign-automation/SKILL.md

@@ -215,15 +215,18 @@ created (draft) -> sent -> delivered -> signed -> completed
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -231,14 +234,16 @@ python3 execution/memory_manager.py store \
   --tags docusign-automation <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/docx/SKILL.md

@@ -204,15 +204,18 @@ Required dependencies (install if not available):
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -220,14 +223,16 @@ python3 execution/memory_manager.py store \
   --tags docx <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/employment-contract-templates/SKILL.md

@@ -46,15 +46,18 @@ Templates and patterns for creating legally sound employment documentation inclu
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -62,14 +65,16 @@ python3 execution/memory_manager.py store \
   --tags employment-contract-templates <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/incident-runbook-templates/SKILL.md

@@ -402,15 +402,18 @@ psql -c "VACUUM FULL large_table;"
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -418,14 +421,16 @@ python3 execution/memory_manager.py store \
   --tags incident-runbook-templates <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/malware-analyst/SKILL.md

@@ -254,15 +254,18 @@ rule Malware_Generic_Packer
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -270,14 +273,16 @@ python3 execution/memory_manager.py store \
   --tags malware-analyst <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/notebooklm/SKILL.md

@@ -276,15 +276,18 @@ Synthesize and respond to user
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -292,14 +295,16 @@ python3 execution/memory_manager.py store \
   --tags notebooklm <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/obsidian-clipper-template-creator/SKILL.md

@@ -71,15 +71,18 @@ See [assets/](assets/) for JSON examples.
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -87,14 +90,16 @@ python3 execution/memory_manager.py store \
   --tags obsidian-clipper-template-creator <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/product-manager-toolkit/SKILL.md

@@ -358,15 +358,18 @@ python scripts/customer_interview_analyzer.py interview.txt json
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -374,14 +377,16 @@ python3 execution/memory_manager.py store \
   --tags product-manager-toolkit <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/readme/SKILL.md

@@ -849,15 +849,18 @@ Write the README directly to `README.md` in the project root.
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -865,14 +868,16 @@ python3 execution/memory_manager.py store \
   --tags readme <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/reference-builder/SKILL.md

@@ -195,15 +195,18 @@ Remember: Your goal is to create reference documentation that answers every poss
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -211,14 +214,16 @@ python3 execution/memory_manager.py store \
   --tags reference-builder <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/research-engineer/SKILL.md

@@ -142,15 +142,18 @@ Using Matrix Exponentiation, we can reduce this to $O(\log n)$. Implementation f
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -158,14 +161,16 @@ python3 execution/memory_manager.py store \
   --tags research-engineer <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/screenshots/SKILL.md

@@ -408,15 +408,18 @@ All screenshots are true retina-quality (2x deviceScaleFactor) and ready for mar
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -424,14 +427,16 @@ python3 execution/memory_manager.py store \
   --tags screenshots <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns

package/templates/skills/extended/documentation/search-specialist/SKILL.md

@@ -87,15 +87,18 @@ Focus on actionable insights. Always provide direct quotes for important claims.
 > **Adapted for [@techwavedev/agi-agent-kit](https://www.npmjs.com/package/@techwavedev/agi-agent-kit)**
 > Original source: [antigravity-awesome-skills](https://github.com/sickn33/antigravity-awesome-skills)
 
-### Qdrant Memory Integration
+### Hybrid Memory Integration (Qdrant + BM25)
 
 Before executing complex tasks with this skill:
 ```bash
 python3 execution/memory_manager.py auto --query "<task summary>"
 ```
+
+**Decision Tree:**
 - **Cache hit?** Use cached response directly — no need to re-process.
 - **Memory match?** Inject `context_chunks` into your reasoning.
 - **No match?** Proceed normally, then store results:
+
 ```bash
 python3 execution/memory_manager.py store \
   --content "Description of what was decided/solved" \
@@ -103,14 +106,16 @@ python3 execution/memory_manager.py store \
   --tags search-specialist <relevant-tags>
 ```
 
+> **Note:** Storing automatically updates both Vector (Qdrant) and Keyword (BM25) indices.
+
 ### Agent Team Collaboration
 
-- This skill can be invoked by the `orchestrator` agent via intelligent routing.
-- In **Agent Teams mode**, results are shared via Qdrant shared memory for cross-agent context.
-- In **Subagent mode**, this skill runs in isolation with its own memory namespace.
+- **Strategy**: This skill communicates via the shared memory system.
+- **Orchestration**: Invoked by `orchestrator` via intelligent routing.
+- **Context Sharing**: Always read previous agent outputs from memory before starting.
 
 ### Local LLM Support
 
 When available, use local Ollama models for embedding and lightweight inference:
 - Embeddings: `nomic-embed-text` via Qdrant memory system
-- Lightweight analysis: Local models reduce API costs for repetitive patterns
+- Lightweight analysis: Local models reduce API costs for repetitive patterns