diff --git a/docs/core_abstraction/node.md b/docs/core_abstraction/node.md index de6586b..bf3360d 100644 --- a/docs/core_abstraction/node.md +++ b/docs/core_abstraction/node.md @@ -10,7 +10,7 @@ nav_order: 1 A **Node** is the smallest building block. Each Node has 3 steps `prep->exec->post`:
- +
1. `prep(shared)` diff --git a/docs/design_pattern/agent.md b/docs/design_pattern/agent.md index 3141bd5..a086426 100644 --- a/docs/design_pattern/agent.md +++ b/docs/design_pattern/agent.md @@ -10,7 +10,7 @@ nav_order: 1 Agent is a powerful design pattern in which nodes can take dynamic actions based on the context.
- +
## Implement Agent with Graph diff --git a/docs/design_pattern/mapreduce.md b/docs/design_pattern/mapreduce.md index de680e3..237b3cb 100644 --- a/docs/design_pattern/mapreduce.md +++ b/docs/design_pattern/mapreduce.md @@ -14,7 +14,7 @@ MapReduce is a design pattern suitable when you have either: and there is a logical way to break the task into smaller, ideally independent parts.
- +
You first break down the task using [BatchNode](../core_abstraction/batch.md) in the map phase, followed by aggregation in the reduce phase. diff --git a/docs/design_pattern/rag.md b/docs/design_pattern/rag.md index a2629e4..a534782 100644 --- a/docs/design_pattern/rag.md +++ b/docs/design_pattern/rag.md @@ -10,7 +10,7 @@ nav_order: 3 For certain LLM tasks like answering questions, providing relevant context is essential. One common architecture is a **two-stage** RAG pipeline:
- +
1. **Offline stage**: Preprocess and index documents ("building the index"). diff --git a/docs/design_pattern/workflow.md b/docs/design_pattern/workflow.md index 476dfb4..92dc536 100644 --- a/docs/design_pattern/workflow.md +++ b/docs/design_pattern/workflow.md @@ -10,7 +10,7 @@ nav_order: 2 Many real-world tasks are too complex for one LLM call. The solution is to **Task Decomposition**: decompose them into a [chain](../core_abstraction/flow.md) of multiple Nodes.
- +
> - You don't want to make each task **too coarse**, because it may be *too complex for one LLM call*. diff --git a/docs/guide.md b/docs/guide.md index 4d29e79..eba0049 100644 --- a/docs/guide.md +++ b/docs/guide.md @@ -56,7 +56,7 @@ Agentic Coding should be a collaboration between Human System Design and Agent I 3. **Utilities**: Based on the Flow Design, identify and implement necessary utility functions. - Think of your AI system as the brain. It needs a body—these *external utility functions*—to interact with the real world: -
+
- Reading inputs (e.g., retrieving Slack messages, reading emails) - Writing outputs (e.g., generating reports, sending emails) @@ -127,7 +127,7 @@ Agentic Coding should be a collaboration between Human System Design and Agent I - > **You'll likely iterate a lot!** Expect to repeat Steps 3–6 hundreds of times. > - >
+ >
{: .best-practice } 8. **Reliability** diff --git a/docs/index.md b/docs/index.md index d16d65a..fa2e0e3 100644 --- a/docs/index.md +++ b/docs/index.md @@ -13,7 +13,7 @@ A [100-line](https://github.com/the-pocket/PocketFlow/blob/main/pocketflow/__ini - **Agentic-Coding**: Intuitive enough for AI agents to help humans build complex LLM applications.
- +
@@ -29,7 +29,7 @@ We model the LLM workflow as a **Graph + Shared Store**: - [(Advanced) Parallel](./core_abstraction/parallel.md) nodes/flows handle I/O-bound tasks.
- +
## Design Pattern @@ -44,7 +44,7 @@ From there, it’s easy to implement popular design patterns: - [(Advanced) Multi-Agents](./design_pattern/multi_agent.md) coordinate multiple agents.
- +
## Utility Function