JobWorkFlow: Recruitment Pipeline

A technical workflow for scouting roles, initializing application artifacts, and tracking pipeline state. The core automation is a Go-based MCP server (kaka-job-scout) plus an Obsidian/Dataview tracker and LaTeX resume templates.

Architecture

1. Scout: External agent/browser finds a job post and extracts structured fields.
2. Initialize: kaka-job-scout creates an Obsidian tracker note and a per-company folder.
3. Tailor: LaTeX template is edited and compiled into a PDF resume.
4. Monitor: Obsidian + Dataview renders the pipeline dashboard.

Requirements

• Go (for mcp-server)
• LaTeX + latexmk (for resume builds)
• Obsidian with Dataview plugin (for tracking)

Quickstart

1) Start the MCP server

cd mcp-server
go run .

2) Initialize a job entry

The MCP tool accepts:

• company (string)
• position (string)
• url (string)
• jd (string)

On success, it writes a tracker note under trackers/ and creates:

data/applications/<company>/resume/
data/applications/<company>/cover/

3) Build a resume PDF

latexmk -pdf -output-directory=data/templates data/templates/resume.tex
latexmk -c -output-directory=data/templates data/templates/resume.tex

4) Track applications

Open trackers/Job Application.md in Obsidian. Add new entries by copying trackers/template.md into a tracker folder and filling frontmatter.

Utility

Check for duplicates before initializing a tracker:

./scripts/check_job.sh "<job_url>" "<Company Name>" "<Position Name>"

Directory Map

• data/templates/ LaTeX sources and template content
• data/applications/ Per-company resume/cover letter output
• trackers/ Obsidian vault and Dataview dashboard
• mcp-server/ Go MCP server (kaka-job-scout)
• docs/skills/ Operational docs for automation/skills
• scripts/ Local helper scripts

Notes

• Build artifacts (*.pdf, *.aux, *.log, *.out, *.synctex.gz) are ignored via .gitignore.