@techwavedev/agi-agent-kit 1.3.5 → 1.4.2

package/templates/skills/extended/ai-agents/nanobanana-ppt-skills/SKILL.md

@@ -29,15 +29,18 @@ For more information, see the [source repository](https://github.com/op7418/Nano
 > **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" \
@@ -45,14 +48,16 @@ python3 execution/memory_manager.py store \
   --tags nanobanana-ppt-skills <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/ai-agents/onboarding-cro/SKILL.md

@@ -440,15 +440,18 @@ If you need more context:
 > **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" \
@@ -456,14 +459,16 @@ python3 execution/memory_manager.py store \
   --tags onboarding-cro <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/ai-agents/paid-ads/SKILL.md

@@ -558,15 +558,18 @@ If you need more context:
 > **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" \
@@ -574,14 +577,16 @@ python3 execution/memory_manager.py store \
   --tags paid-ads <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/ai-agents/performance-testing-review-ai-review/SKILL.md

@@ -457,15 +457,18 @@ Use this tool to transform code review from manual process to automated AI-assis
 > **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" \
@@ -473,14 +476,16 @@ python3 execution/memory_manager.py store \
   --tags performance-testing-review-ai-review <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/ai-agents/plaid-fintech/SKILL.md

@@ -57,15 +57,18 @@ Listen for PENDING_DISCONNECT webhook to proactively prompt users.
 > **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" \
@@ -73,14 +76,16 @@ python3 execution/memory_manager.py store \
   --tags plaid-fintech <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/ai-agents/prompt-caching/SKILL.md

@@ -68,15 +68,18 @@ Works well with: `context-window-management`, `rag-implementation`, `conversatio
 > **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" \
@@ -84,14 +87,16 @@ python3 execution/memory_manager.py store \
   --tags prompt-caching <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/ai-agents/prompt-engineer/SKILL.md

@@ -259,15 +259,18 @@ The skill is entirely self-contained, operating purely on user input and framewo
 > **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" \
@@ -275,14 +278,16 @@ python3 execution/memory_manager.py store \
   --tags prompt-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/ai-agents/prompt-engineering/SKILL.md

@@ -178,15 +178,18 @@ Build prompts that gracefully handle failures:
 > **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" \
@@ -194,14 +197,16 @@ python3 execution/memory_manager.py store \
   --tags prompt-engineering <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/ai-agents/prompt-library/SKILL.md

@@ -329,15 +329,18 @@ When crafting prompts, ensure:
 > **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" \
@@ -345,14 +348,16 @@ python3 execution/memory_manager.py store \
   --tags prompt-library <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/ai-agents/rag-engineer/SKILL.md

@@ -97,15 +97,18 @@ Works well with: `ai-agents-architect`, `prompt-engineer`, `database-architect`,
 > **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" \
@@ -113,14 +116,16 @@ python3 execution/memory_manager.py store \
   --tags rag-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/ai-agents/rag-implementation/SKILL.md

@@ -428,15 +428,18 @@ def evaluate_rag_system(qa_chain, test_cases):
 > **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" \
@@ -444,14 +447,16 @@ python3 execution/memory_manager.py store \
   --tags rag-implementation <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/ai-agents/sales-automator/SKILL.md

@@ -62,15 +62,18 @@ Write conversationally. Show empathy for customer problems.
 > **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" \
@@ -78,14 +81,16 @@ python3 execution/memory_manager.py store \
   --tags sales-automator <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/ai-agents/shodan-reconnaissance/SKILL.md

@@ -510,30 +510,35 @@ shodan download network_scan.json.gz 'net:192.168.1.0/24'
 > **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" \
   --type decision \
-  --tags shodan-reconnaissance <relevant-tags>
+  --tags Shodan Reconnaissance and Pentesting <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/ai-agents/skill-developer/SKILL.md

@@ -433,15 +433,18 @@ See [TROUBLESHOOTING.md](TROUBLESHOOTING.md) for complete debugging guide.
 > **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" \
@@ -449,14 +452,16 @@ python3 execution/memory_manager.py store \
   --tags skill-developer <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/ai-agents/skill-rails-upgrade/SKILL.md

@@ -415,15 +415,18 @@ Use this skill when working with analyze rails apps and provide upgrade assessme
 > **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" \
@@ -431,14 +434,16 @@ python3 execution/memory_manager.py store \
   --tags skill-rails-upgrade <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