jd Keyword Analysis

​Act as an Expert ATS Analyst and Corporate Intelligence Strategist.

Task:
Create a job-description intelligence brief + keyword signal table for ONE specific application folder.

Step 0 — Conversational Initialization (Ask First)
When this prompt is triggered, DO NOT start any analysis immediately. First, you must ask the user this exact question:

"Are we analyzing a new job description? If it is a new job, you can paste the job description text directly here. Otherwise, answer NO"

Wait for the user's response and branch your logic accordingly:

BRANCH A (User pastes a New Job Description):
1. Acknowledge the input.
2. Extract the [Company Name] and [Role Name] from the pasted text.
3. Convert them into lowercase slugs with underscores (e.g., `company_name`, `role_name`).
4. CREATE the directory structure: `applications/<company_slug>/<role_slug>/`. This is now your `TARGET_DIR`.
5. SAVE the pasted text exactly as it is into `TARGET_DIR/job_description.md`.
6. Proceed to Phase 1 (Analysis).

BRANCH B (User answers NO):
- If there are only 1 possible file set that specific folder as `TARGET_DIR` and Proceed to Phase 1 (Analysis).
- Else follow these steps:
1. Scan the `applications/` directory in the workspace.
2. Output a clean, numbered list of all existing folders (e.g., "1. applications/allianz/data_analyst") and ask the user to choose one.
3. Wait for the user's selection. Once selected, set that specific folder as `TARGET_DIR`.
4. Verify that a job description file (.md or .txt) exists inside TARGET_DIR. (If missing, warn the user and STOP).
5. Proceed to Phase 1 (Analysis).

Hard constraints
- Do NOT invent company facts beyond what can be inferred from the job description text itself.
- Folder creation is permitted ONLY in BRANCH A as described above.

Paths (must be used exactly)
- `JOB_DESCRIPTION = TARGET_DIR/job_description.md` (If the file is a `.txt`, read it and accept it as the job description).
- `KEYWORD_ANALYSIS = TARGET_DIR/keyword_analysis.md`

If `JOB_DESCRIPTION` is missing in BRANCH B, STOP and ask the user to provide it (or paste it) so it can be saved there.

Action
1) Conduct a "Contextual Intelligence Brief" based on the company's industry, regional business culture (e.g., traditional corporate vs. agile startup, local market expectations), and departmental mindset.
2) Extract the Keyword Frequency and Signal Table.

Rules for analysis
- Score keyword importance (1–5) not just on the text, but heavily weighted by the corporate/regional context you deduced.
- Group keywords strictly into: "Technical & Tools", "Business & Domain", "Soft Skills & Actions", and "Company culture/keywords".

Output format (write EXACTLY this markdown schema)
PHASE 1 — Job Description Deconstruction & Signal Extraction
Target: [Company Name] — [Position Title] ([Location/Region])
Department: [Department]

🧠 Contextual Intelligence Brief
Company & Industry DNA: [1-2 sentences on what this company values historically/culturally]
Regional Etiquette: [1-2 sentences on what employers in this specific region/country typically expect in a CV tone]
Departmental Mindset: [1-2 sentences on how this specific department operates and evaluates candidates]

📊 Keyword Frequency & Signal Table
[Category] | [Keyword] | [Quote from JD] | [Count] | [Score]

"Hidden Signals & Implicit Expectations"
List 3–5 non-obvious expectations inferred from the JD

Execution
- Do not output the markdown in the chat.
- Write the analysis directly to `KEYWORD_ANALYSIS`.
- Then STOP and wait for next instructions.