kyle.belanger/ai_cover_letter_generator
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

provide tailored Resume

Browse source
This commit is contained in:
Kyle Belanger 2025-02-20 09:19:20 -05:00
parent f12c949bc8
commit 099ccc40e8
6 changed files with 40 additions and 9 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
data/resume_parser_api.json
View file

@ -8,7 +8,7 @@
"content": [
{
"type": "text",
"text": "You are an advanced resume parsing system designed to extract key information from resumes and organize it into a structured format. Your task is to analyze the following resume and create a JSON object containing the most relevant and insightful information.\n\nHere is the resume you need to parse:\n\n<resume>\n{{resume}}\n</resume>\n\nPlease follow these steps to complete the task:\n\n1. Carefully read through the entire resume.\n\n2. Extract the following information:\n - First name\n - Last name\n - Location\n - Work experience\n - School experience\n - Skills\n\n3. For each category, focus on the most insightful and relevant information.\n\n4. Organize the extracted information into a JSON object with the following keys:\n - first_name\n - last_name\n - location\n - work_experience\n - school_experience\n - skills\n\nBefore providing the final JSON output, break down the resume inside <resume_breakdown> tags:\n\n- Quote the relevant sections for first name, last name, and location.\n- For work experience, list out each entry with numbers, quoting relevant details for company, position, duration, and key responsibilities.\n- For school experience, list out each entry with numbers, quoting relevant details for institution, degree, field of study, and graduation year.\n- For skills, categorize them into technical and soft skills, listing each skill.\n- Double-check that all required fields in the JSON structure have corresponding information from the resume.\n\nThis breakdown will help ensure a thorough and accurate parsing of the resume.\n\nHere's an example of the expected JSON structure (note that this is just a structural example and does not contain actual data):\n\n```json\n{\n \"first_name\": \"String\",\n \"last_name\": \"String\",\n \"location\": \"String\",\n \"work_experience\": [\n {\n \"company\": \"String\",\n \"position\": \"String\",\n \"duration\": \"String\",\n \"responsibilities\": [\"String\", \"String\"]\n }\n ],\n \"school_experience\": [\n {\n \"institution\": \"String\",\n \"degree\": \"String\",\n \"field_of_study\": \"String\",\n \"graduation_year\": \"Number\"\n }\n ],\n \"skills\": [\"String\", \"String\", \"String\"]\n}\n```\n\nPlease ensure that your final JSON output follows this structure, but with actual data extracted from the provided resume."
"text": "Here is the resume you need to parse:\n\n<resume>\n{{resume}}\n</resume>\n\nYou are an advanced resume parsing system designed to extract key information from resumes and organize it into a structured JSON format. Your task is to analyze the above resume and create a JSON object containing the most relevant and insightful information.\n\nPlease follow these steps to complete the task:\n\n1. Carefully read through the entire resume.\n\n2. Extract the following information:\n - First name\n - Last name\n - Location\n - Email\n - Phone Number\n - Work experience\n - School experience\n - Skills\n\n3. For each category, focus on the most insightful and relevant information.\n\n4. Wrap the resume breakdown in <resume_analysis> tags:\n\n - Quote the relevant sections for first name, last name, email, phone number, and location.\n - For work experience, list out each entry with numbers, quoting relevant details for the company, position, duration, and key responsibilities. Note any achievements or quantifiable results.\n - For school experience, list out each entry with numbers, quoting relevant details for institution, degree, field of study, and graduation year. Include any notable academic achievements.\n - For skills, categorize them into technical skills, soft skills, and domain-specific skills, listing each skill.\n - Identify any unique selling points or standout information in the resume.\n - If any information is missing from the resume, explicitly state that it's not found.\n\n5. Organize the extracted information into a JSON object with the following structure:\n\n```json\n{\n \"first_name\": \"String\",\n \"last_name\": \"String\",\n \"location\": \"String\",\n \"email\": \"String\",\n \"phone_number\": \"String\",\n \"work_experience\": [\n {\n \"company\": \"String\",\n \"position\": \"String\",\n \"duration\": \"String\",\n \"responsibilities\": [\"String\", \"String\"],\n \"achievements\": [\"String\", \"String\"]\n }\n ],\n \"school_experience\": [\n {\n \"institution\": \"String\",\n \"degree\": \"String\",\n \"field_of_study\": \"String\",\n \"graduation_year\": \"Number\",\n \"achievements\": [\"String\", \"String\"]\n }\n ],\n \"skills\": {\n \"technical\": [\"String\", \"String\"],\n \"soft\": [\"String\", \"String\"],\n \"domain_specific\": [\"String\", \"String\"]\n },\n \"unique_selling_points\": [\"String\", \"String\"]\n}\n```\n\nImportant: \n- Include all specified categories in the JSON output, even if the information is not found in the resume. \n- For missing information, use empty string values (\"\") or empty arrays ([]) as appropriate.\n- For work_experience and school_experience, if no entries are found, include an empty array.\n- For skills, if no skills are found in a category, include an empty array for that category.\n\nAfter the <resume_analysis> section, provide the final JSON output in <json_output> tags. Ensure that your JSON output follows the structure provided above, with actual data extracted from the resume or empty values for missing information."
}
]
},

7
public/css/base_styles.css
View file

@ -11,8 +11,13 @@
display: none;
}
#tailoredResumeSection {
display: none;
}
@layer base {
textarea {
resize: auto;
}
}
}

3
public/css/styles.css
View file

@ -549,6 +549,9 @@
#coverLetterSection {
display: none;
}
#tailoredResumeSection {
display: none;
}
@layer base {
textarea {
resize: auto;

6
public/index.html
View file

@ -34,6 +34,12 @@
<textarea id="coverLetterOutput" rows="15" cols="80"></textarea>
<button id="downloadBtn">Download as DOCX</button>
</div>
<div id="tailoredResumeSection">
<h3>Tailored Resume:</h3>
<textarea id="tailoredResumeOutput" rows="15" cols="80"></textarea>
<h3>Key Updates</h3>
<pre id="keyResumeUpdates"></pre>
</div>
<!-- End Hidden Section -->

3
public/js/script.js
View file

@ -104,6 +104,9 @@ document.getElementById("generateCoverLetterBtn").addEventListener("click", asyn
const formattedText = formatCoverLetter(data.coverLetter);
document.getElementById("coverLetterOutput").value = formattedText;
document.getElementById("coverLetterSection").style.display = "grid"; // Show cover letter section
document.getElementById("tailoredResumeSection").style.display = "grid";
document.getElementById("tailoredResumeOutput").value = data.tailoredResumeText;
document.getElementById("keyResumeUpdates").textContent = data.resumeChanges;
generateBtn.textContent = "Generate New Cover Letter"
}
} catch (error) {

28
routes/generate.js
View file

@ -48,7 +48,7 @@ router.post('/extract-resume', upload.single('resume'), async (req, res) => {
const resumeResponse = await anthropic.messages.create(resume_parser_api);
// Extract JSON-formatted profile from the response
const candidateProfile = resumeResponse.content[0].text.split('```json')[1].split('```')[0].trim();
const candidateProfile = resumeResponse.content[0].text.split('<json_output>')[1].split('</json_output>')[0].trim();
res.json({candidateProfile: JSON.parse(candidateProfile) });
@ -61,20 +61,34 @@ router.post('/extract-resume', upload.single('resume'), async (req, res) => {
// Handle Resume upload and user input for Job Description
router.post('/', async (req, res) => {
try {
// Load Variables
const { candidateProfile, jobDescription, keyPoints } = req.body;
const coverLetterAPI = require('../data/cover_letter_api.json');
const resumeTailoringAPI = require('../data/resume_tailoring_api.json');
const cover_letter_api = require('../data/cover_letter_api.json');
cover_letter_api.messages[0].content[0].text = cover_letter_api.messages[0].content[0].text
// Replace API placeholders
coverLetterAPI.messages[0].content[0].text = coverLetterAPI.messages[0].content[0].text
.replace('{{resume_json}}', candidateProfile)
.replace('{{job_description}}', jobDescription)
.replace('{{key_points}}', keyPoints )
.replace('{{date}}', new Date().toDateString());
const coverLetterResponse = await anthropic.messages.create(cover_letter_api);
const coverLetterRawText = coverLetterResponse.content[0].text.split('<cover_letter>')[1].split('</cover_letter>')[0].trim();
res.json({ coverLetter: coverLetterRawText });
resumeTailoringAPI.messages[0].content[0].text = resumeTailoringAPI.messages[0].content[0].text
.replace('{{resume_json}}', candidateProfile)
.replace('{{job_description}}', jobDescription)
// Send Messages to LLM
const coverLetterResponse = await anthropic.messages.create(coverLetterAPI);
const tailoredResumeResponse = await anthropic.messages.create(resumeTailoringAPI);
// Extract Relevant Responses
const coverLetterRawText = coverLetterResponse.content[0].text.split('<cover_letter>')[1].split('</cover_letter>')[0].trim();
const tailoredResumeText = tailoredResumeResponse.content[0].text.split('<tailored_resume>')[1].split('</tailored_resume>')[0].trim();
const resumeChanges = tailoredResumeResponse.content[0].text.split('<key_changes>')[1].split('</key_changes>')[0].trim();
res.json({ coverLetter: coverLetterRawText, tailoredResumeText, resumeChanges });
} catch (error) {
console.error(error);
res.status(500).json({ error: 'Error generating cover letter' });