Classify Items with 302.AI API on New Session Created from Zep API

Pipedream makes it easy to connect APIs for 302.AI, Zep and 2,900+ other apps remarkably fast.

Trigger workflow on

New Session Created from the Zep API

Next, do this

Classify Items with the 302.AI API

No credit card required

▶

Watch us build a workflow

8 min

Watch now ➜

Trusted by 1,000,000+ developers from startups to Fortune 500 companies

Developers ♥ Pipedream

Getting Started#

This integration creates a workflow with a Zep trigger and 302.AI action. When you configure and deploy the workflow, it will run on Pipedream's servers 24x7 for free.

Select this integration
Configure the New Session Created trigger
1. Connect your Zep account
2. Configure timer
Configure the Classify Items action
1. Connect your 302.AI account
2. Select a Model
3. Optional- Configure Max Tokens
4. Optional- Configure Temperature
5. Optional- Configure Top P
6. Optional- Configure N
7. Optional- Configure Stop
8. Optional- Configure Presence Penalty
9. Optional- Configure Frequency Penalty
10. Optional- Configure User
11. Configure info
12. Configure Items
13. Configure Categories
Deploy the workflow
Send a test event to validate your setup
Turn on the trigger

Details#

This integration uses pre-built, source-available components from Pipedream's GitHub repo. These components are developed by Pipedream and the community, and verified and maintained by Pipedream.

To contribute an update to an existing component or create a new component, create a PR on GitHub. If you're new to Pipedream component development, you can start with quickstarts for trigger span and action development, and then review the component API reference.

Trigger#

New Session Created on Zep

Description:Emit new event when a new session is created in Zep. [See the documentation](https://help.getzep.com/api-reference/memory/list-sessions)

Version:0.0.3

Key:zep-new-session

View on GitHub

Trigger Code#

import common from "../common/base.mjs";

export default {
  ...common,
  key: "zep-new-session",
  name: "New Session Created",
  description: "Emit new event when a new session is created in Zep. [See the documentation](https://help.getzep.com/api-reference/memory/list-sessions)",
  version: "0.0.3",
  type: "source",
  dedupe: "unique",
  methods: {
    ...common.methods,
    getNewResults(lastTs, max) {
      return this.getSessions({
        lastTs,
        max,
        orderBy: "created_at",
      });
    },
    generateMeta(session) {
      return {
        id: session.session_id,
        summary: `New Session: ${session.session_id}`,
        ts: Date.parse(session.created_at),
      };
    },
  },
};

Trigger Configuration#

This component may be configured based on the props defined in the component code. Pipedream automatically prompts for input values in the UI and CLI.

Label	Prop	Type	Description
Zep	`zep`	`app`	This component uses the Zep app.
N/A	`db`	`$.service.db`	This component uses `$.service.db` to maintain state between executions.
	`timer`	`$.interface.timer`

Trigger Authentication#

Zep uses API keys for authentication. When you connect your Zep account, Pipedream securely stores the keys so you can easily authenticate to Zep APIs in both code and no-code steps.

About Zep#

The Foundational Memory Layer for AI. Equip your agents with the knowledge to complete tasks, from the mundane to monumental.

Action#

Classify Items on 302.AI

Description:Classify input items into predefined categories using 302.AI models. Perfect for tagging, segmentation, and automated organization. [See documentation](https://doc.302.ai/147522039e0)

Version:0.0.1

Key:_302_ai-classify-items

View on GitHub

Action Code#

import common from "../common/common-helper.mjs";

export default {
  ...common,
  name: "Classify Items",
  version: "0.0.1",
  annotations: {
    destructiveHint: false,
    openWorldHint: true,
    readOnlyHint: true,
  },
  key: "_302_ai-classify-items",
  description: "Classify input items into predefined categories using 302.AI models. Perfect for tagging, segmentation, and automated organization. [See documentation](https://doc.302.ai/147522039e0)",
  type: "action",
  props: {
    ...common.props,
    info: {
      type: "alert",
      alertType: "info",
      content: `Provide a list of **items** and a list of **categories**. The output will contain an array of objects, each with properties \`item\` and \`category\`
        \nExample:
        \nIf **Categories** is \`["people", "pets"]\`, and **Items** is \`["dog", "George Washington"]\`
        \n The output will contain the following categorizations:
        \n \`[{"item":"dog","category":"pets"},{"item":"George Washington","category":"people"}]\`
      `,
    },
    items: {
      label: "Items",
      description: "Items to categorize",
      type: "string[]",
    },
    categories: {
      label: "Categories",
      description: "Categories to classify items into",
      type: "string[]",
    },
  },
  methods: {
    ...common.methods,
    systemInstructions() {
      return "Your goal is to categorize items into specific categories and produce ONLY JSON. The user will provide both the items and categories. Please only categorize items into the specific categories, and no others, and output ONLY JSON.";
    },
    outputFormat() {
      return "Please only categorize items into the specific categories, and no others. Output a valid JSON string — an array of objects, where each object has the following properties: item, category. Do not return any English text other than the JSON, either before or after the JSON. I need to parse the response as JSON, and parsing will fail if you return any English before or after the JSON";
    },
    userMessage() {
      return `Categorize each of the following items:\n\n${this.items.join("\n")}\n\ninto one of the following categories:\n\n${this.categories.join("\n")}\n\n${this.outputFormat()}`;
    },
    summary() {
      return `Categorized ${this.items.length} items into ${this.categories.length} categories`;
    },
    formatOutput({
      messages, response,
    }) {
      if (!messages || !response) {
        throw new Error("Invalid API output, please reach out to https://pipedream.com/support");
      }
      const responses = response.choices?.map(({ message }) => message.content);
      const categorizations = [];
      for (const response of responses) {
        try {
          categorizations.push(JSON.parse(response));
        } catch (err) {
          console.log("Failed to parse output, assistant returned malformed JSON");
        }
      }
      const output = {
        messages,
      };
      const n = this.n
        ? parseInt(this.n)
        : 1;
      if (n > 1) {
        output.categorizations = categorizations;
      } else {
        output.categorizations = categorizations[0];
      }
      return output;
    },
  },
};

Action Configuration#

This component may be configured based on the props defined in the component code. Pipedream automatically prompts for input values in the UI.

Label	Prop	Type	Description
302.AI	`_302_ai`	`app`	This component uses the 302.AI app.
Model	`modelId`	`string`	Select a value from the drop down menu.
Max Tokens	`maxTokens`	`string`	The maximum number of tokens to generate in the completion.
Temperature	`temperature`	`string`	What sampling temperature to use, between 0 and 2. Higher values like 0.8 will make the output more random, while lower values like 0.2 will make it more focused and deterministic.
Top P	`topP`	`string`	An alternative to sampling with temperature, called nucleus sampling, where the model considers the results of the tokens with top_p probability mass. So 0.1 means only the tokens comprising the top 10% probability mass are considered.
N	`n`	`string`	How many completions to generate for each prompt
Stop	`stop`	`string[]`	Up to 4 sequences where the API will stop generating further tokens. The returned text will not contain the stop sequence.
Presence Penalty	`presencePenalty`	`string`	Number between -2.0 and 2.0. Positive values penalize new tokens based on whether they appear in the text so far, increasing the model's likelihood to talk about new topics.
Frequency Penalty	`frequencyPenalty`	`string`	Number between -2.0 and 2.0. Positive values penalize new tokens based on their existing frequency in the text so far, decreasing the model's likelihood to repeat the same line verbatim.
User	`user`	`string`	A unique identifier representing your end-user, which can help monitor and detect abuse.
Items	`items`	`string[]`	Items to categorize
Categories	`categories`	`string[]`	Categories to classify items into