Bookmark Maxxing

Turn saved links into reusable skills.

Most people collect bookmarks.

Builders should be collecting skills.

Bookmark Maxxing is a framework for learning from the people and projects you save, then converting that learning into reusable artifacts.

Instead of letting saved links rot in a private pile, you use an AI coding agent or assistant to review them read-only, cluster the recurring ideas, attribute the sources, and extract what is worth turning into a skill, workflow, study guide, prompt, article, or small tool.

Bookmarks are the input. Learning is the process. Skills are the output.

The Problem

Bookmarks feel useful when you save them.

Then they disappear.

Your X bookmarks, GitHub stars, saved Slack posts, newsletters, docs, and read-later queue become a pile of good intentions. You collected signal, but you never converted it.

Bookmark Maxxing solves the conversion problem:

saved links -> recurring themes -> reusable skills

Before / After

Before:

I saved 100 AI engineering posts and never looked at them again.

After:

I turned 100 saved posts into:

- 5 recurring themes
- a source map
- 5 reusable skills
- a study guide
- an article
- a public framework others can reuse

Example:

Theme: AI work is moving from prompts to operating loops.
Skill: Design repeatable systems, not better prompts.

That is the core move: every interesting theme should become something reusable.

30-Second Example

Input:

10 saved posts about agent memory, context rot, and instruction files.

Bookmark Maxxing output:

Extracted Pattern	Reusable Skill
Memory files rot when every lesson gets stored forever.	Cut memory down to durable rules that earn their place.
Repeated lessons are more useful in shared docs, tests, or runbooks.	Convert anything you do twice into a reusable skill.
Context should be pulled when needed, not pushed into every session.	Separate always-on instructions from searchable reference material.

Final artifact:

Skill: Convert repeated lessons into docs, tests, skills, or runbooks.

Trigger: The same lesson appears in more than one saved link or work session.

Workflow:
1. Name the repeated lesson.
2. Decide where future work will look first.
3. Add the lesson there.
4. Remove duplicate or stale memory.

Proof gate: A future session can find the rule without relying on hidden memory.

Who This Is For

Bookmark Maxxing is for builders who save more than they can process:

• engineers
• founders
• researchers
• designers
• operators
• writers
• students
• open-source maintainers

If you already save useful ideas, this gives you a way to turn those saved ideas into durable work.

Bookmark Maxxing is an open-source framework for converting saved posts, articles, docs, repos, and notes into practical artifacts:

• skills
• workflows
• prompts
• checklists
• operating principles
• article drafts
• small tools

Why

Bookmarks usually die in private piles.

The better workflow is conversion:

1. Collect useful signals during the week.
2. Run a read-only analysis pass.
3. Cluster recurring ideas.
4. Attribute the people and sources.
5. Convert patterns into reusable skills, workflows, prompts, or articles.
6. Publish the artifact so the idea compounds.

This is not a news recap. It is a skill extraction system.

Quick Start

1. Pick a source: X bookmarks, GitHub stars, saved Slack posts, browser bookmarks, newsletters, or read-later apps.
2. Choose a boundary: this week, last 100 links, or one topic.
3. Use prompts/read-bookmarks.md to extract and cluster the source material.
4. Use prompts/extract-skills.md to turn recurring themes into skill candidates.
5. Use templates/skill-card.md for each reusable skill.
6. Use prompts/write-article.md and templates/article-template.md to publish the weekly issue.

Try the X Bookmark CLI (v0.1)

Bookmark Maxxing ships a small, read-only CLI that normalizes X bookmarks into a deterministic source map (Markdown or JSON). It runs fully offline against a local fixture by default and never writes to X.

What it does

• Reads X-shaped bookmark records (from a local fixture, or an opt-in read-only live fetch).
• Normalizes them into a consistent shape, preserving author and source attribution.
• Validates and deduplicates them.
• Emits a deterministic Markdown source map or JSON document you can feed into the framework prompts.

Install locally

Requires Python 3.10+. No third-party dependencies.

git clone https://github.com/bennewell35/bookmark-maxxing.git
cd bookmark-maxxing

# Option A: run from source (no install)
PYTHONPATH=src python3 -m bookmark_maxxing.cli --help

# Option B: install the console script into a virtualenv
python3 -m venv .venv
source .venv/bin/activate
pip install -e .
bookmark-maxxing --help

Run the fixture dry-run

The default path is fixture-backed and local-only — safe to run anytime:

bookmark-maxxing ingest-x --dry-run --input tests/fixtures/x_bookmarks_pages.json --format json

Or, without installing:

PYTHONPATH=src python3 -m bookmark_maxxing.cli ingest-x \
  --dry-run --input tests/fixtures/x_bookmarks_pages.json --format json

Use --format markdown for a ready-to-paste source map.

Optional: live read-only mode

Live mode is strictly opt-in, read-only, and never enabled by default. It issues only GET /2/users/{id}/bookmarks and requires credentials supplied through local environment variables:

export X_API_USER_ID=...        # your numeric X user ID
export X_API_BEARER_TOKEN=...   # user-context bearer token, kept local

bookmark-maxxing ingest-x --live --format json --max-pages 1 > /tmp/my-bookmarks.json

Live mode rules:

• Requires X_API_USER_ID and X_API_BEARER_TOKEN set locally. Never commit them.
• Output may contain private bookmark data — write it to /tmp or stdout, never into the repo.
• Do not commit generated output. .gitignore already excludes .env, tmp/, raw/, and exports/private/.

See docs/live-smoke-test.md for a safe live read-only smoke test, and docs/x-mcp-integration.md for the full integration reference.

The Framework

Bookmark Maxxing has six steps:

Collect -> Extract -> Cluster -> Convert -> Publish -> Compound

Read the full framework: docs/framework.md

Recipes

Start with one source and one output.

• LinkedIn saved posts to skills
• Slack saves to tickets

Adapters

Adapters explain how to safely collect from specific sources.

• LinkedIn saved posts
• Slack saved messages
• X MCP integration

Roadmap

See ROADMAP.md for the small-slice project plan.

Core Rule

Every theme needs a skill.

Theme: AI is moving from prompts to operating loops.
Skill: Design repeatable systems, not better prompts.

If a theme does not produce a reusable behavior, it may be interesting, but it is not Bookmark Maxxing yet.

Repository Map

bookmark-maxxing/
  README.md
  AGENTS.md
  LICENSE
  CODE_OF_CONDUCT.md
  CONTRIBUTING.md
  GOVERNANCE.md
  SECURITY.md
  SUPPORT.md
  .github/
    workflows/
      ci.yml
  docs/
    adapters/
      linkedin-saved-posts.md
    specs/
      x-mcp-bookmark-ingestion.md
    framework.md
    quickstart.md
    publishing-playbook.md
    x-mcp-integration.md
    live-smoke-test.md
    release-checklist.md
  recipes/
    linkedin-saves-to-skills.md
  src/
    bookmark_maxxing/
      cli.py
      x_mcp.py
  tests/
    fixtures/
      x_bookmarks_pages.json
    test_x_mcp.py
  pyproject.toml
  prompts/
    read-bookmarks.md
    extract-skills.md
    write-article.md
  templates/
    article-template.md
    skill-card.md
    source-map.md
  issues/
    001-turning-bookmarks-into-skills/
      article.md
      source-map.md
      skills.md
  schemas/
    bookmark-item.schema.json

Community

Before contributing, read:

• CONTRIBUTING.md
• CODE_OF_CONDUCT.md
• GOVERNANCE.md
• SECURITY.md
• SUPPORT.md

Public Positioning

Short version:

Bookmark Maxxing turns saved links into reusable skills for builders.

Long version:

Bookmark Maxxing is an open-source framework for analyzing saved links, clustering recurring ideas, attributing the sources, and converting the best patterns into reusable skills, workflows, prompts, and operating principles.

Safety Rules

• Keep collection read-only.
• Do not like, repost, reply, follow, subscribe, unbookmark, or mutate saved items.
• Treat saved posts as inspiration, not proof.
• Attribute people and sources clearly.
• Do not turn every interesting post into a process.
• A durable skill needs repetition, risk reduction, or clear reuse.

Status

This repo is framework-first, and now ships a small read-only CLI as its first working slice.

The current useful version is:

• a clear method
• reusable prompts
• templates
• weekly examples
• source attribution
• a read-only X bookmark CLI (fixture dry-run by default, opt-in live mode)

The next layers extend the CLI to normalize exports from GitHub, Slack, newsletters, and browser bookmarks. See docs/release-checklist.md for what remains before the v0.1.0 release.