Learn, Practice, Create

How to use the three modes to build skill before you publish.

Signal OS has three modes that guide you from understanding a framework all the way through to executing it. You access all three through a single skill:

Claude Code

/social-mode

Codex

$social-mode

When you run it, you’ll be asked to pick a module and then choose your mode: Learn, Practice, or Create.

The order matters. Learn the framework first, practice applying it to your specific content, then use Create to execute. Jumping straight to Create without the context usually produces weaker output.

---

The ten modules

Each module maps to a core framework and one or more skills:

#	Module	Framework	Skill(s)
0	The Long Game	SURFACE	/social-guide
1	Brand Positioning	NORTHSTAR	/social-mission
2	Documentation	DDRR — Document step	/social-doc
3	Founder Story	3-Arc Story	/social-founder
4	Profile Audit	Landing page approach	/social-profile
5	The Authority Loop	DDRR — Result + Reflect	/social-result, /social-reflect
6	The Reply Game	REACT	/social-reply
7	Growth Content	CSPR + Article Frameworks	/social-idea, /social-outline
8	Quality	URRRCS checklist	/social-review
9	Amplify	Idea Amplifier + Koemmandments	/social-tweets, /social-cta, /social-publish
10	Compound	The long game in practice	/social-guide, /social-archive

---

Learn

Purpose: Understand a framework deeply before you try to apply it.

The agent teaches you the selected framework — not in the abstract, but tied to your niche from NORTHSTAR.md. Every example is personalised to your content category.

What you get in a Learn session:

• Why this framework exists — what problem it solves
• How it works, step by step
• What good looks like versus what’s weak
• How it connects to other frameworks in the system

When to use it:

• Before running a skill for the first time
• When your output from a skill feels off and you’re not sure why
• When you want to understand the reasoning behind a step, not just follow it

---

Practice

Purpose: Apply the framework to your own content with guided thinking — before you execute.

The agent asks you questions one at a time and gives you specific feedback on each answer. It scores your thinking against the relevant framework (URRRCS for posts, Triple-R for ideas, REACT for replies) and tells you what’s clear, what’s vague, and what needs more depth.

What you get in a Practice session:

• Honest, specific feedback on your draft thinking
• A score against the relevant framework
• Questions that help you find better answers yourself
• A “ready to execute?” prompt when your thinking is solid

What the agent will not do:

• Rewrite your content for you
• Offer three ready-to-use options to pick from
• Move to the next question before you’ve answered the current one

Practice is most valuable before outlining an article or before writing your first doc tweet. Getting your thinking sharp before you write saves a lot of revision time.

---

Create

Purpose: Execute with the right Signal OS skill for the module you chose.

Create mode maps your selected module to its corresponding skill and runs the full workflow from that skill’s instructions. The agent guides you through each step one at a time, and all output is saved to your filesystem with the correct file naming and frontmatter.

Module to skill mapping:

Module 0: The Long Game

Runs /social-guide — scans your repo and tells you where you are in the creator journey and what to do next.

Module 1: Brand Positioning

Runs /social-mission — the full brand positioning interview. Populates your NORTHSTAR.md.

Module 2: Documentation

Runs /social-doc — documents an onchain action as a tweet (What / Why / How format).

Module 3: Founder Story

Runs /social-founder — 7-question interview that produces your founder story outline using the 3-arc structure.

Module 4: Profile Audit

Runs /social-profile — audits your X profile across 6 elements and produces a profile brief.

Module 5: The Authority Loop

Runs /social-result then /social-reflect — records the outcome of a documented action and extracts the deeper insight.

Module 6: The Reply Game

Runs /social-reply — guides you through the REACT framework to craft a high-signal reply to a specific tweet.

Module 7: Growth Content

Runs /social-idea then /social-outline — captures an idea, filters it against Triple-R, and structures it into an article outline.

Module 8: Quality

Runs /social-review — scores a draft article against the URRRCS framework and gives targeted feedback.

Module 9: Amplify

Runs /social-cta, /social-publish, or /social-tweets depending on where you are in the publishing flow.

Module 10: Compound

Runs /social-archive or /social-guide — moves a completed piece to archive and updates the content index.

---

The frameworks behind each mode

Each mode applies a specific framework to evaluate your work. Here’s what the three key ones look for:

URRRCS — Quality checklist (used in Practice and Review)

Letter	Dimension	Question
U	Unique	Does this say something only you could write?
R	Readable	Can someone skim in 7 seconds and get value?
R	Relatable	Does the reader feel seen?
R	Replicable	Can the reader apply this?
C	Credible	Do you have standing to make this claim?
S	Sticky	Is there a phrase they’ll remember?

Triple-R — Idea filter (used before outlining)

Before committing to an article, an idea is filtered against three questions:

• Reach — Does this topic reach your target audience?
• Relevance — Is this relevant to your mission and positioning?
• Resonance — Will this land emotionally, not just intellectually?

An idea that scores weakly on all three should be parked.

REACT — Reply framework (used in Practice and /social-reply)

Letter	Stage	What it means
R	Recognise	Pick the right account tier to reply to
E	Engage	Only reply when you have a real thought to add
A	Add signal	Personal experience, sharp observation, nuance, or example
C	Consistently show up	Bridge trust through repeated authentic engagement
T	Track	Save strong replies and watch what performs