Implementar function_call para traer el plan de estudios #8

New Issue

Open

opened 2026-01-28 15:25:58 +00:00 by alexrg · 2 comments

alexrg commented 2026-01-28 15:25:58 +00:00

Owner

Copy Link

Copy Source

Entrada

function proponer_mejora(plan_id, input, campose) {

const plan_id = '4660b774-a6e2-44f4-a340-747b41235c7c';

let input = [
  { role: "user", content: `ID del plan de estudio: ${plan_id}` },
  { role: "user", content: "Mejora el nombre del plan de estudio" },
];

const campos = ['nombre']

Código

const {error, data} = await supabase.from('planes_estudio').select('estructuras_plan (definicion), conversation_id').eq('id', plan_id).single();
if (error) {
  console.error("Error al obtener la definición del plan de estudio:", error);
}
const definicion = data?.estructuras_plan?.definicion;
definicion.properties = Object.fromEntries(campos.map(key => [key, definicion.properties[key]]));
definicion.required = definicion.required.filter((key: string) => campos.includes(key));

const conversation_id = data?.conversation_id;

// 1. Define a list of callable tools for the model
const tools = [
  {
    type: "function",
    name: "obtener_plan_estudio",
    description: "Obtener el plan de estudio por ID.",
    parameters: {
      type: "object",
      properties: {
        id: {
          type: "string",
          description: "El ID del plan de estudio.",
        },
      },
      required: ["id"],
    },
  },
];


// 2. Prompt the model with tools defined
let response = await client.responses.create({
  model: "gpt-5-nano",
  tools,
  input,
});

input.push(...response.output);

// Check if the model decided to call any tool
for (const item of response.output.filter(item => item.type === 'function_call') || []) {
  switch(item.name) {
    case "obtener_plan_estudio":
      // 3. Execute the function logic for obtener_plan_estudio
      const args = JSON.parse(item.arguments);
      const { data, error } = await supabase
        .from('planes_estudio')
        .select('datos')
        .eq('id', args.id)
        .single();

      if (error) {
        throw new Error(`Error al obtener el plan de estudio: ${error.message}`);
      }
      const planEstudio = data?.datos;
    // 4. Add the tool response to the input for the model
    input.push({
      type: "function_call_output",
      call_id: item.call_id,
      output: JSON.stringify(planEstudio)
    });
  }
}

response = await client.responses.create({
  model: "gpt-5-nano",
  instructions: "Propón mejoras solo con la información del plan de estudio obtenido por la herramienta.",
  tools,
  input,
  conversation: conversation_id,
  reasoning: {effort: "high"},

  text: {format: { type: "json_schema", name: "PlanEstudioMejorado", schema: definicion }},
});

console.log("Salida final:", response.output);
}

Entrada ```js function proponer_mejora(plan_id, input, campose) { const plan_id = '4660b774-a6e2-44f4-a340-747b41235c7c'; let input = [ { role: "user", content: `ID del plan de estudio: ${plan_id}` }, { role: "user", content: "Mejora el nombre del plan de estudio" }, ]; const campos = ['nombre'] ``` Código ```ts const {error, data} = await supabase.from('planes_estudio').select('estructuras_plan (definicion), conversation_id').eq('id', plan_id).single(); if (error) { console.error("Error al obtener la definición del plan de estudio:", error); } const definicion = data?.estructuras_plan?.definicion; definicion.properties = Object.fromEntries(campos.map(key => [key, definicion.properties[key]])); definicion.required = definicion.required.filter((key: string) => campos.includes(key)); const conversation_id = data?.conversation_id; // 1. Define a list of callable tools for the model const tools = [ { type: "function", name: "obtener_plan_estudio", description: "Obtener el plan de estudio por ID.", parameters: { type: "object", properties: { id: { type: "string", description: "El ID del plan de estudio.", }, }, required: ["id"], }, }, ]; // 2. Prompt the model with tools defined let response = await client.responses.create({ model: "gpt-5-nano", tools, input, }); input.push(...response.output); // Check if the model decided to call any tool for (const item of response.output.filter(item => item.type === 'function_call') || []) { switch(item.name) { case "obtener_plan_estudio": // 3. Execute the function logic for obtener_plan_estudio const args = JSON.parse(item.arguments); const { data, error } = await supabase .from('planes_estudio') .select('datos') .eq('id', args.id) .single(); if (error) { throw new Error(`Error al obtener el plan de estudio: ${error.message}`); } const planEstudio = data?.datos; // 4. Add the tool response to the input for the model input.push({ type: "function_call_output", call_id: item.call_id, output: JSON.stringify(planEstudio) }); } } response = await client.responses.create({ model: "gpt-5-nano", instructions: "Propón mejoras solo con la información del plan de estudio obtenido por la herramienta.", tools, input, conversation: conversation_id, reasoning: {effort: "high"}, text: {format: { type: "json_schema", name: "PlanEstudioMejorado", schema: definicion }}, }); console.log("Salida final:", response.output); } ```

Guillermo.Arrieta was assigned by alexrg 2026-01-28 15:25:59 +00:00

alexrg commented 2026-01-28 17:29:17 +00:00

Author

Owner

Copy Link

Copy Source

Ejemplo de OPENAI

// 1. Define a list of callable tools for the model
const tools = [
  {
    type: "function",
    name: "get_horoscope",
    description: "Get today's horoscope for an astrological sign.",
    parameters: {
      type: "object",
      properties: {
        sign: {
          type: "string",
          description: "An astrological sign like Taurus or Aquarius",
        },
      },
      required: ["sign"],
    },
  },
];

function getHoroscope(sign: string) {
  return sign + " Next Tuesday you will befriend a baby otter.";
}

// Create a running input list we will add to over time
let input = [
  { role: "user", content: "What is my horoscope? I am an Aquarius." },
];

// 2. Prompt the model with tools defined
let response = await client.responses.create({
  model: "gpt-5-nano",
  tools,
  input,
});

input.push(...response.output);

// Check if the model decided to call any tool
response.output?.forEach((item) => {
  if (item.type === "function_call" && item.name === "get_horoscope") {
    // 3. Execute the function logic for get_horoscope
    const args = JSON.parse(item.arguments);
    const horoscope = getHoroscope(args.sign);
    
    // 4. Add the tool response to the input for the model
    input.push({
      type: "function_call_output",
      call_id: item.call_id,
      output: JSON.stringify({ horoscope })
    });
  }
});

console.log(input);


response = await client.responses.create({
  model: "gpt-5-nano",
  instructions: "Respond only with a horoscope generated by a tool.",
  tools,
  input,
});

// 5. The model should be able to give a response!
console.log("Final output:");
console.log(response.output);

# Ejemplo de OPENAI ```ts // 1. Define a list of callable tools for the model const tools = [ { type: "function", name: "get_horoscope", description: "Get today's horoscope for an astrological sign.", parameters: { type: "object", properties: { sign: { type: "string", description: "An astrological sign like Taurus or Aquarius", }, }, required: ["sign"], }, }, ]; function getHoroscope(sign: string) { return sign + " Next Tuesday you will befriend a baby otter."; } // Create a running input list we will add to over time let input = [ { role: "user", content: "What is my horoscope? I am an Aquarius." }, ]; // 2. Prompt the model with tools defined let response = await client.responses.create({ model: "gpt-5-nano", tools, input, }); input.push(...response.output); // Check if the model decided to call any tool response.output?.forEach((item) => { if (item.type === "function_call" && item.name === "get_horoscope") { // 3. Execute the function logic for get_horoscope const args = JSON.parse(item.arguments); const horoscope = getHoroscope(args.sign); // 4. Add the tool response to the input for the model input.push({ type: "function_call_output", call_id: item.call_id, output: JSON.stringify({ horoscope }) }); } }); console.log(input); response = await client.responses.create({ model: "gpt-5-nano", instructions: "Respond only with a horoscope generated by a tool.", tools, input, }); // 5. The model should be able to give a response! console.log("Final output:"); console.log(response.output); ```

alexrg commented 2026-01-28 20:14:37 +00:00

Author

Owner

Copy Link

Copy Source

Ejemplo en bruto

const plan_id = '4660b774-a6e2-44f4-a340-747b41235c7c';

let input = [
  { role: "user", content: `ID del asignatura: ${plan_id}` },
  { role: "user", content: "Mejora el nombre del asignatura" },
];

const campos = ['nombre']

// 1. Define a list of callable tools for the model
const tools = [
  {
    type: "function",
    name: "obtener_plan_estudio",
    description: "Obtener el plan de estudio por ID.",
    parameters: {
      type: "object",
      properties: {
        id: {
          type: "string",
          description: "El ID del plan de estudio.",
        },
      },
      required: ["id"],
    },
  },
  {
    type: "function",
    name: "obtener_asignatura",
    description: "Obtener la asignatura por ID.",
    parameters: {
      type: "object",
      properties: {
        id: {
          type: "string",
          description: "El ID del asignatura.",
        },
      },
      required: ["id"],
    },
  },
];


// 2. Prompt the model with tools defined
let response = await client.responses.create({
  model: "gpt-5-nano",
  tools,
  input,
});

input.push(...response.output);
let conversation_id = null, definicion = null;
// Check if the model decided to call any tool
for (const item of response.output.filter(item => item.type === 'function_call') || []) {
  switch(item.name) {
    case "obtener_plan_estudio":
    const {error, data} = await supabase.from('planes_estudio').select('estructuras_plan (definicion), conversation_id').eq('id', plan_id).single();
    if (error) {
    console.error("Error al obtener la definición del plan de estudio:", error);
    }
    definicion = data?.estructuras_plan?.definicion;
    conversation_id = data?.conversation_id;

          // 3. Execute the function logic for obtener_plan_estudio
          const args = JSON.parse(item.arguments);
          const { data, error } = await supabase
            .from('planes_estudio')
            .select('datos')
            .eq('id', args.id)
            .single();

          if (error) {
            throw new Error(`Error al obtener el plan de estudio: ${error.message}`);
          }
          const planEstudio = data?.datos;
        // 4. Add the tool response to the input for the model
        input.push({
          type: "function_call_output",
          call_id: item.call_id,
          output: JSON.stringify(planEstudio)
        });
      }
    case "obtener_plan_estudio":
    const {error, data} = await supabase.from('planes_estudio').select('estructuras_asignatura (definicion), conversation_id').eq('id', plan_id).single();
    if (error) {
    console.error("Error al obtener la definición del plan de estudio:", error);
    }
    definicion = data?.estructuras_asignatura?.definicion;
    conversation_id = data?.conversation_id;

          // 3. Execute the function logic for obtener_plan_estudio
          const args = JSON.parse(item.arguments);
          const { data, error } = await supabase
            .from('planes_estudio')
            .select('datos')
            .eq('id', args.id)
            .single();

          if (error) {
            throw new Error(`Error al obtener el plan de estudio: ${error.message}`);
          }
          const planEstudio = data?.datos;
        // 4. Add the tool response to the input for the model
        input.push({
          type: "function_call_output",
          call_id: item.call_id,
          output: JSON.stringify(planEstudio)
        });
      }
}
definicion.properties = Object.fromEntries(campos.map(key => [key, definicion.properties[key]]));
definicion.required = definicion.required.filter((key: string) => campos.includes(key));

response = await client.responses.create({
  model: "gpt-5-nano",
  instructions: "Propón mejoras solo con la información del plan de estudio obtenido por la herramienta.",
  tools,
  input,
  conversation: conversation_id,
  reasoning: {effort: "high"},

  text: {format: { type: "json_schema", name: "PlanEstudioMejorado", schema: definicion }},
});

console.log("Salida final:", response.output);

# Ejemplo en bruto ```js const plan_id = '4660b774-a6e2-44f4-a340-747b41235c7c'; let input = [ { role: "user", content: `ID del asignatura: ${plan_id}` }, { role: "user", content: "Mejora el nombre del asignatura" }, ]; const campos = ['nombre'] // 1. Define a list of callable tools for the model const tools = [ { type: "function", name: "obtener_plan_estudio", description: "Obtener el plan de estudio por ID.", parameters: { type: "object", properties: { id: { type: "string", description: "El ID del plan de estudio.", }, }, required: ["id"], }, }, { type: "function", name: "obtener_asignatura", description: "Obtener la asignatura por ID.", parameters: { type: "object", properties: { id: { type: "string", description: "El ID del asignatura.", }, }, required: ["id"], }, }, ]; // 2. Prompt the model with tools defined let response = await client.responses.create({ model: "gpt-5-nano", tools, input, }); input.push(...response.output); let conversation_id = null, definicion = null; // Check if the model decided to call any tool for (const item of response.output.filter(item => item.type === 'function_call') || []) { switch(item.name) { case "obtener_plan_estudio": const {error, data} = await supabase.from('planes_estudio').select('estructuras_plan (definicion), conversation_id').eq('id', plan_id).single(); if (error) { console.error("Error al obtener la definición del plan de estudio:", error); } definicion = data?.estructuras_plan?.definicion; conversation_id = data?.conversation_id; // 3. Execute the function logic for obtener_plan_estudio const args = JSON.parse(item.arguments); const { data, error } = await supabase .from('planes_estudio') .select('datos') .eq('id', args.id) .single(); if (error) { throw new Error(`Error al obtener el plan de estudio: ${error.message}`); } const planEstudio = data?.datos; // 4. Add the tool response to the input for the model input.push({ type: "function_call_output", call_id: item.call_id, output: JSON.stringify(planEstudio) }); } case "obtener_plan_estudio": const {error, data} = await supabase.from('planes_estudio').select('estructuras_asignatura (definicion), conversation_id').eq('id', plan_id).single(); if (error) { console.error("Error al obtener la definición del plan de estudio:", error); } definicion = data?.estructuras_asignatura?.definicion; conversation_id = data?.conversation_id; // 3. Execute the function logic for obtener_plan_estudio const args = JSON.parse(item.arguments); const { data, error } = await supabase .from('planes_estudio') .select('datos') .eq('id', args.id) .single(); if (error) { throw new Error(`Error al obtener el plan de estudio: ${error.message}`); } const planEstudio = data?.datos; // 4. Add the tool response to the input for the model input.push({ type: "function_call_output", call_id: item.call_id, output: JSON.stringify(planEstudio) }); } } definicion.properties = Object.fromEntries(campos.map(key => [key, definicion.properties[key]])); definicion.required = definicion.required.filter((key: string) => campos.includes(key)); response = await client.responses.create({ model: "gpt-5-nano", instructions: "Propón mejoras solo con la información del plan de estudio obtenido por la herramienta.", tools, input, conversation: conversation_id, reasoning: {effort: "high"}, text: {format: { type: "json_schema", name: "PlanEstudioMejorado", schema: definicion }}, }); console.log("Salida final:", response.output); ```

Sign in to join this conversation.

No Branch/Tag Specified

Branches Tags

main

No results found.

Labels

Clear labels

Compat/Breaking

Breaking change that won't be backward compatible

Kind/Bug

Something is not working

Kind/Documentation

Documentation changes

Kind/Enhancement

Improve existing functionality

Kind/Feature

New functionality

Kind/Security

This is security issue

Kind/Testing

Issue or pull request related to testing

Priority

Critical

1

The priority is critical

Priority

High

2

The priority is high

Priority

Low

4

The priority is low

Priority

Medium

3

The priority is medium

Reviewed

Confirmed

1

Issue has been confirmed

Reviewed

Duplicate

2

This issue or pull request already exists

Reviewed

Invalid

3

Invalid issue

Reviewed

Won't Fix

3

This issue won't be fixed

Status

Abandoned

3

Somebody has started to work on this but abandoned work

Status

Blocked

1

Something is blocking this issue or pull request

Status

Need More Info

2

Feedback is required to reproduce issue or to continue work

No labels

Milestone

No items

No Milestone

Assignees

Clear assignees

Guillermo.Arrieta roberto.silva (Roberto Silva Corona) alexrg (Alejandro Rosales)

No Assignees Guillermo.Arrieta

1 Participants

Notifications

Subscribe

Due Date

No due date set.

Reference: alexrg/genesis-2#8