agentleFS
slowbutfastslowbutfast/agentic-resume-builder/AGENTS.md

agentic-resume-builder

AGENTS.md0 starsChanged 4 months agoMITView on GitHub

Welcome! If you are an AI coding assistant (Cursor, Anti-Gravity CLI, OpenCode, Claude Code, OpenChamber, or Gemini CLI) helping a user generate or edit their resume using this repository, follow these core directives. Content & Density Rules: - First Bullet Rule: Bullet #1 under any project/role MUST clearly explain what was built and what it does in plain, accessible English. Save deep technical mechanisms for Bullets #2 and #3. - Selective Bolding: Bold quantitative metrics (e.g. \textbf{97% reduction}, \textbf{5+ hours/week})…

  • Where files are headingNew AGENTS.md files have more rules about keeping secrets out of the code, and fewer about code style.

    Of the AGENTS.md files we read, the 33,315 first seen in the last six months, against the older ones: the kind of rule that gained most share, and the kind that lost most. About files like this in general, not this one, and an estimate, since a kind is a pattern in the words.

  • In this repository3 other files it gives its agents.

The file

Show its rules and text

We split this AGENTS.md into the rules it gives an agent

14 rules. About 0 are common, 5 are distinctive, and 9 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 9

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    When a user asks you to explain or set up their resume:

    line 23
  2. Very specific:

    Check if data/resume_bank.json exists.

    line 24
  3. Distinctive:

    Run schema validation to verify environment:

    line 28
  4. Very specific:

    Run python3 build_resume.py --summary (or -s) for a clean overview tree of active roles, entry keys, and bullet IDs.

    line 34
  5. Very specific:

    Run python3 build_resume.py --help for full CLI documentation and example workflows.

    line 35
  6. Very specific:

    Run python3 build_resume.py --schema to view JSON structure rules.

    line 36
  7. Very specific:

    Use deterministic CLI CRUD flags to modify data/resume_bank.json safely without manual formatting errors:

    line 39
  8. Distinctive:

    Line Length Target: Keep bullet text around 220–245 characters so each bullet spans exactly 2 full lines edge-to-edge across ~85%–95% of …

    line 47
  9. Distinctive:

    No Orphan Widows: Avoid single-word trailing lines (<4 trailing words).

    line 48
  10. Very specific:

    LaTeX Escaping: Do NOT manually add LaTeX backslashes for &, %, $, #, _ inside JSON strings.

    line 50
  11. Distinctive:

    Run single-role build or build all:

    line 53
  12. Very specific:

    Ensure all variants report Pages: 1 and PASSED (1 Page).

    line 67
  13. Very specific:

    If OVERFLOW warning occurs: trim 3–5 words from longer bullets using python3 build_resume.py --edit-bullet <ID> --text "...".

    line 68
  14. Very specific:

    If UNDERFLOW warning occurs: expand bullets or add a project key to target role.

    line 69

≈ “Common” is an estimate: at least 1 in 5 of the 42,283 AGENTS.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
AGENTS.md

AI Agent Directives for agentic-resume-builder

Welcome! If you are an AI coding assistant (Cursor, Anti-Gravity CLI, OpenCode, Claude Code, OpenChamber, or Gemini CLI) helping a user generate or edit their resume using this repository, follow these core directives.


🎯 System Architecture Overview

  1. Master Source of Truth:
    • Primary data bank: data/resume_bank.json (created by copying data/resume_bank.example.json).
    • Schema specification: data/resume_bank.schema.json.
  2. LaTeX Baseline Template:
    • Located in templates/main.tex (based on Jake Gutierrez's Gold-Standard Resume Template).
  3. Compilation Pipeline:
    • Engine script: src/generate_resumes.py.
    • CLI entry point: build_resume.py.

🤖 How to Help the User (Step-by-Step Agent Workflow)

Step 1: Initial Setup Check

When a user asks you to explain or set up their resume:

  1. Check if data/resume_bank.json exists. If not, instruct or run:
    cp data/resume_bank.example.json data/resume_bank.json
  2. Run schema validation to verify environment:
    python3 build_resume.py --lint

Step 2: Querying Resume Roles, Slugs & Structure

  • Run python3 build_resume.py --summary (or -s) for a clean overview tree of active roles, entry keys, and bullet IDs.
  • Run python3 build_resume.py --help for full CLI documentation and example workflows.
  • Run python3 build_resume.py --schema to view JSON structure rules.

Step 3: CLI CRUD Operations & Editing Heuristics

Use deterministic CLI CRUD flags to modify data/resume_bank.json safely without manual formatting errors:

  • Role CRUD: --add-role, --edit-role, --delete-role
  • Entry CRUD: --add-project, --add-experience, --delete-entry
  • Bullet CRUD: --add-bullet, --edit-bullet, --delete-bullet

Content & Density Rules:

  • First Bullet Rule: Bullet #1 under any project/role MUST clearly explain what was built and what it does in plain, accessible English. Save deep technical mechanisms for Bullets #2 and #3.
  • Selective Bolding: Bold quantitative metrics (e.g. \textbf{97% reduction}, \textbf{5+ hours/week}) and core tech stack names to draw recruiter visual focus. Do NOT bold full sentences.
  • Line Length Target: Keep bullet text around 220–245 characters so each bullet spans exactly 2 full lines edge-to-edge across ~85%–95% of printable line width.
  • No Orphan Widows: Avoid single-word trailing lines (<4 trailing words). Either trim 2–3 words or expand by 10–15 characters to fill the line out.
  • Date Consistency: Format all dates consistently using abbreviated months (e.g. Jun 2026 – Present, Aug 2025 – May 2026).
  • LaTeX Escaping: Do NOT manually add LaTeX backslashes for &, %, $, #, _ inside JSON strings. The python builder handles escaping automatically.

Step 4: Compiling & Verifying

Run single-role build or build all:

python3 build_resume.py --role swe
python3 build_resume.py

Build a shareable copy with personal contact info masked (phone, email, linkedin, github by default):

python3 build_resume.py --role swe --sanitize
python3 build_resume.py --role swe --sanitize phone,email
# Accepts a comma-separated subset of name,phone,email,linkedin,github.
# PDF is saved as <role>_sanitized.pdf so the real resume is never overwritten.

Check stdout metrics:

  • Ensure all variants report Pages: 1 and PASSED (1 Page).
  • If OVERFLOW warning occurs: trim 3–5 words from longer bullets using python3 build_resume.py --edit-bullet <ID> --text "...".
  • If UNDERFLOW warning occurs: expand bullets or add a project key to target role.

🛠️ Specialized Agent Skills Directory

Refer to dedicated instruction files in the skills/ directory for detailed specialized agent loops:

  • skills/project-auditor/SKILL.md: Technical Resume Auditor prompt to inspect external codebases and extract raw bullet drafts.
  • skills/resume-bank-editor/SKILL.md: Autonomous agent loop for role selection, bullet density tuning, building, and page budget verification.
  • skills/resume-optimizer/SKILL.md: Single-page height density optimizer guide.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.