56roadmap entries
54published builds
8research blocks
3public repos
A living index of NosisTech agent engineering builds. 56 documented patterns tracked openly as the R&D roadmap evolves.
Current Snapshot
- 56 roadmap entries: 54 published educational builds and
- Published: live article available on nosistech.com.
- Skipped: intentionally removed from the active build sequence, but retained in the index for transparency.
- Future additions: new related builds may be added beyond the original 50+ roadmap as the research library grows.
Agent Patterns and Builds
| # | Build / Pattern | Focus | Block | Status |
|---|---|---|---|---|
| 1 | Tool Invocation Agent | Tool invocation and real-world actions | Block 1 | Published |
| 2 | Autonomous Decision-Making Agent | Autonomous triage and escalation rules | Block 1 | Published |
| 3 | Market Data Agent | Market data lookup and summarization | Block 1 | Published |
| 4 | Financial News Agent | Financial news briefing | Block 1 | Published |
| 5 | Async Standup Summarizer | Team update summarization | Block 1 | Published |
| 6 | Flowise | No-code agent workflow | Block 1 | Published |
| 7 | PraisonAI | Multi-agent research workflow | Block 1 | Published |
| 8 | Memory-Augmented Agent | Persistent memory | Block 2 | Published |
| 9 | Knowledge Retrieval Agent (RAG) | Retrieval-augmented generation | Block 2 | Published |
| 10 | Document Intelligence Agent | OCR, extraction, and review routing | Block 2 | Published |
| 11 | Letta (MemGPT) | Long-term memory manager | Block 2 | Published |
| 12 | AnythingLLM | Private document AI workspace | Block 2 | Published |
| 13 | Phidata | Memory, knowledge, and tool use | Block 2 | Published |
| 14 | Planning Agent | Plan generation and execution | Block 3 | Published |
| 15 | Data Analysis Agent | Statistical analysis plus narrative | Block 3 | Published |
| 16 | Marketing Content Assistant | Brand-compliant content workflow | Block 3 | Published |
| 17 | Physical World Sensing Agent | Sensor simulation and action policy | Block 3 | Published |
| 18 | Langflow | Visual AI flow builder | Block 3 | Published |
| 19 | DSPy | Prediction and prompt optimization | Block 3 | Published |
| 20 | Chain-of-Agents Orchestrator | Sequential specialist agents | Block 4 | Published |
| 21 | Conflict Resolution Agent | Dual-review conflict escalation | Block 4 | Published |
| 22 | Financial Advisory Agent | Market and news advisory briefing | Block 4 | Published |
| 23 | CrewAI | Research crew orchestration | Block 4 | Published |
| 24 | AutoGen | Agent collaboration patterns | Block 4 | Published |
| 25 | Agency Swarm | Visible agent orchestration | Block 4 | Published |
| 26 | SuperAGI | Multi-agent dashboard framework | Block 4 | Skipped |
| 27 | Ethical Reasoning Agent | High-stakes reasoning pattern | Block 5 | Skipped |
| 28 | Explainable Agent (LIME/SHAP) | Explainability and confidence | Block 5 | Published |
| 29 | Compliance-Driven Agent | Rule-bound AI code workflow | Block 5 | Published |
| 30 | AI System Governance Intake Agent | Governance intake and review flags | Block 5 | Published |
| 31 | LLM-Guard | AI safety gateway | Block 5 | Published |
| 32 | Griptape | Governance pipeline | Block 5 | Published |
| 33 | Pydantic-AI | Structured governance decisions | Block 5 | Published |
| 34 | Fabric | Pattern library prompt runner | Block 6 | Published |
| 35 | Prompt Injection Defense Agent | Prompt injection screening | Block 6 | Published |
| 36 | Hallucination Detection Agent | Unsupported-claim detection | Block 6 | Published |
| 37 | Verification & Validation Agent | Claim checking against trusted data | Block 6 | Published |
| 38 | Promptfoo | Red-team evaluation harness | Block 6 | Published |
| 39 | PentestGPT | Authorized review planning | Block 6 | Published |
| 40 | GDPR / CCPA Compliance Agent | Data privacy compliance pattern | Block 6 | Published |
| 41 | Browser-Use | Web task planning | Block 7 | Published |
| 42 | GPT-Researcher | Research synthesis agent | Block 7 | Published |
| 43 | Goose | Action planning workflow | Block 7 | Published |
| 44 | OpenHands | Sandbox planning agent | Block 7 | Published |
| 45 | SWE-agent | Software engineering next-action planner | Block 7 | Published |
| 46 | TaskWeaver | Stateful data analysis workflow | Block 7 | Published |
| 47 | Self-Improving Agent | Self-improvement with guardrails | Block 8 | Published |
| 48 | General Problem Solver | General reasoning workflow | Block 8 | Published |
| 49 | EvoAgentX | Framework rebuild | Block 8 | Published |
| 50 | MetaGPT | Simulated software company | Block 8 | Published |
| 51 | SmolAgents | Code agent pattern | Block 8 | Published |
| 52 | Bayesian Operations Gateway | Bayesian-style risk scoring and escalation thresholds | Block 8 | Published |
| 53 | Knowledge Gap Discovery Agent | Discovery and ranking of product knowledge gaps | Block 8 | Published |
| 54 | Collective Intelligence Consensus Engine | Multi-agent consensus with adversarial review | Block 8 | Published |
| 55 | Aegis Constraint Agent | Deterministic constraint gates with model-assisted explanation | Block 8 | Published |
| 56 | The Flipped Interaction Pattern: Agent Asks First | Agent-driven context gathering before task execution | Block 9 | Published |
Related Operating Notes
These notes sit alongside the numbered build roadmap and cover practical operating lessons that shape AI system cost, reliability, governance, and deployment decisions.