docs(docs): add initial project documentation and README files

AGENTS.md

@@ -0,0 +1,19 @@
+---
+framework_version: 1.0.0
+---
+
+# Agent Guidelines: AI Job Search
+
+This workspace is structured to manage job search activities, scraper tools, CVs, cover letters, and interview preparation.
+
+## Thin-Pointer Design (Single Source of Truth)
+
+To prevent duplication and configuration drift across different AI agent frameworks (Claude Code, Google Antigravity, Codex, Cursor, Gemini CLI, etc.), this workspace uses a unified thin-pointer design. All agent runtimes should load the canonical specifications and candidate profiles from the files and directories below:
+
+1. **Personal Candidate Profile:**
+   - The candidate profile, contact details, education, and target preferences are defined in [CLAUDE.md](CLAUDE.md) and the individual profile methodology files under [.claude/skills/job-application-assistant/](.claude/skills/job-application-assistant/) (specifically `01-*.md` etc.).
+2. **Canonical Workflow Specifications:**
+   - The step-by-step instructions and triggers for tasks (setup, scrape, rank, apply, upskill, interview) are defined in the [.claude/](.claude/) directory (specifically under `.claude/skills/` and `.claude/commands/`).
+   - Do not duplicate these rules or specifications. Treat `.claude/` files as the single source of truth.
+3. **Portal Search Skills:**
+   - Job-portal search CLIs live under [.agents/skills/](.agents/skills/) in the portable Agent Skills format (with a `SKILL.md` per portal). Codex and Antigravity discover these automatically; the `/scrape` workflow in [.claude/skills/job-scraper/](.claude/skills/job-scraper/) orchestrates them.