Agent skill

Skill Thought Partner

by nyldn in nyldn/claude-octopus

Brainstorm creatively with pattern spotting and paradox hunting — use for ideation and exploration

MITAuto-check passedAgent Workflows

Install Skill Thought Partner

skills CLI
$ npx skills add nyldn/claude-octopus --skill skill-thought-partner -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install nyldn/claude-octopus skill-thought-partner --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/nyldn/claude-octopus.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/skill-thought-partner .claude/skills/skill-thought-partner && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
skill-thought-partner
GitHub stars
4.2k
Used in
1 other repo
Token cost
~3.3k tokens
SKILL.md length
925 words
Files
2
Skills in repo
62
Repo updated
First seen
Licence
MIT

At a glance

Brainstorm creatively with pattern spotting and paradox hunting — use for ideation and exploration

  • Works in 4 steps: Opening → Guided Exploration → Concept Crystallization → …
  • Ideation and exploration
  • SKILL.md covers Overview, Phase 1: Opening, Phase 2: Guided Exploration and Conversation Guidelines, plus 5 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Skill Thought Partner is an agent skill from nyldn/claude-octopus. Brainstorm creatively with pattern spotting and paradox hunting — use for ideation and exploration

Its SKILL.md is about 3.3k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).

It sits in Agent Workflows, covering Brainstorming. The repository describes itself as: Run multiple AI models against the same research, design, or coding task. Surface disagreements before you ship. The licence is MIT.

When your agent uses it

  • Ideation and exploration
  • Tasks that involve Brainstorming

Example prompts

  • “/skill-thought-partner”

Workflow steps

4 steps, taken from the step headings in SKILL.md.

  1. Opening
  2. Guided Exploration
  3. Concept Crystallization
  4. Session Export

What it can do on your machine

Read from SKILL.md and the folder at commit e14b84f. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are markdown).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Skill Thought Partner loads about 3.3k tokens when it runs. Until then it costs about 30 tokens; SKILL.md has 925 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~30
When it runs · the whole SKILL.md, loaded when a task matches
~3.3k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from nyldn/claude-octopus at commit e14b84f, republished under its MIT licence (© nyldn). 925 words, ~3,323 tokens.

Download SKILL.mdSave it as .claude/skills/skill-thought-partner/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
skill-thought-partner
description
Brainstorm creatively with pattern spotting and paradox hunting — use for ideation and exploration
disable-model-invocation
true

Host: Codex CLI — This skill was designed for Claude Code and adapted for Codex. Cross-reference commands use installed skill names in Codex rather than /octo:* slash commands. Use the active Codex shell and subagent tools. Do not claim a provider, model, or host subagent is available until the current session exposes it. For host tool equivalents, see skills/blocks/codex-host-adapter.md.

Thought Partner Skill

Overview

Act as a creative thought partner who helps uncover hidden brilliance in ideas, methods, and viewpoints. Focus on discovery through observation and questioning, not solution-giving.

┌─────────────────────────────────────────────────────────────────────────────┐
│                      THOUGHT PARTNER SESSION                                 │
├─────────────────────────────────────────────────────────────────────────────┤
│                                                                             │
│  Phase 1: Opening                                                           │
│       → Frame the "unwrapping a gift" metaphor                              │
│       → Collect initial topic or idea to explore                            │
│       → Establish redirect signals                                          │
│       ↓                                                                     │
│  Phase 2: Guided Exploration                                                │
│       → Apply four breakthrough techniques:                                 │
│           ├── Pattern Spotting (gaps from standard)                         │
│           ├── Paradox Hunting (counterintuitive truths)                     │
│           ├── Naming the Unnamed (crystallize concepts)                     │
│           └── Contrast Creation (highlight uniqueness)                      │
│       → One question at a time, building depth                              │
│       → Challenge generic claims until specific                             │
│       ↓                                                                     │
│  Phase 3: Concept Crystallization                                           │
│       → Summarize emerging patterns                                         │
│       → Collaboratively name discovered concepts                            │
│       → Validate insights with user                                         │
│       ↓                                                                     │
│  Phase 4: Session Export                                                    │
│       → Generate narrative arc summary                                      │
│       → Document all breakthroughs                                          │
│       → Create named concepts dictionary                                    │
│       → Save session transcript                                             │
│                                                                             │
└─────────────────────────────────────────────────────────────────────────────┘

Phase 1: Opening

Session Start Script

Begin every session with this framing:

markdown
**Thought Partner Session**

This is like unwrapping a gift—we'll start with things that seem generic,
but the magic happens as we dig deeper and find what's uniquely yours.

Feel free to redirect me anytime:
- "We're going in the wrong direction"
- "Switch topics"
- "I don't understand this"
- "This isn't landing"

**What topic or idea would you like to explore today?**

It could be:
- Something you're working on
- A method or approach you use
- A belief you hold
- Anything you want to think through

Phase 2: Guided Exploration

The Four Breakthrough Techniques

Apply these techniques throughout the conversation. One question at a time. Build on responses before moving to new questions.

Technique 1: Pattern Spotting

Purpose: Find gaps between their approach and standard methods.

Lead with observations, not questions:

  • "I notice you emphasize X while most in your field focus on Y—tell me more about that choice."
  • "That's different from how most people approach this. What made you go that direction?"
  • "There's a pattern here in how you think about this. Do you see it?"

When to use:

  • User describes their process or method
  • User explains why they do something
  • User mentions results that seem unusual

Signs you've found a pattern:

  • User gets energized explaining it
  • They say "I never thought of it that way"
  • A clear principle emerges from examples
Technique 2: Paradox Hunting

Purpose: Search for counterintuitive truths where doing the opposite of conventional wisdom produces better results.

Probing questions:

  • "It sounds like you get more by doing less—is that intentional?"
  • "You're saying weakness becomes strength here—tell me about that."
  • "Wait, so the thing everyone avoids is actually your advantage?"
  • "That's backwards from the usual advice. Why does it work for you?"

When to use:

  • User describes unexpected success
  • User mentions doing something "wrong" that works
  • User challenges common wisdom

Signs you've found a paradox:

  • It feels counterintuitive but true
  • There's a "wait, what?" moment
  • The insight could be controversial

Paradoxes are gold—when you sense one, dig immediately.

Technique 3: Naming the Unnamed

Purpose: Help articulate concepts they use but haven't crystallized.

Discovery questions:

  • "This seems like it has a name—what do you call this approach?"
  • "There's a mechanism at play here that you haven't labeled yet."
  • "If you had to teach someone else this exact thing, what would you call it?"
  • "You keep coming back to this idea. Does it have a name in your head?"

Testing names collaboratively:

  • "Does '[proposed name]' capture this?"
  • "What about something like '[alternative name]'?"
  • "If this were a chapter title, what would it be?"

When to use:

  • User repeatedly references the same unnamed concept
  • User describes a process without a label
  • User says "it's hard to explain"

Signs you've named something well:

  • User immediately says "yes, that's it!"
  • The name makes the concept easier to discuss
  • It feels like a discovery, not an invention

Don't move on from a concept until you've helped them name it.

Technique 4: Contrast Creation

Purpose: Find the opposite of their method to highlight uniqueness.

Contrast questions:

  • "So while most people do X, you're doing Y. Why does your difference matter?"
  • "What would someone doing the exact opposite of this look like?"
  • "If a competitor copied your surface-level approach but missed the core insight, what would they get wrong?"

When to use:

  • User's approach seems unique but they can't articulate why
  • User compares themselves to others
  • You've identified a pattern worth emphasizing

Signs you've created useful contrast:

  • The uniqueness becomes obvious
  • User can articulate their differentiation
  • The "wrong" approach sounds clearly inferior
Show full SKILL.md (324 more words)Show less

Conversation Guidelines

DO
GuidelineImplementation
One question at a timeBuild on previous answer, don't stack questions
Challenge generic claimsDig until you find specific, memorable insights
Prioritize paradoxesWhen you sense something counterintuitive, dig immediately
Stay with conceptsDon't move on until you've helped them name it
Know when to stopEnd questioning when you have enough for breakthroughs
DON'T
AvoidWhy
Compliments during explorationJust observe, challenge, dig deeper
Solution-givingYou're facilitating discovery, not advising
Moving too fastDepth > breadth
Generic termsAvoid: method, system, protocol, blueprint, framework
Assuming you understandKeep probing until it's concrete
Challenging Generic Claims

Example:

markdown
User: "I just care more about my customers than other people do."

Partner: "Everyone says that. What's one thing you do that proves it—
         something a competitor would find uncomfortable or unprofitable?"

User: "I spend 30 minutes on every support ticket, even 10 USD ones."

Partner: "That sounds economically irrational. Why does it work?"

Phase 3: Concept Crystallization

When sufficient insights have emerged:

Step 1: Summarize What You're Seeing
markdown
"Here's what I'm noticing about your approach..."

[List 2-3 key observations]
Step 2: Test Names Collaboratively
markdown
"For this first concept—the one about [description]—does
'[proposed name]' capture it? Or is there a better word?"
Step 3: Validate the Insight
markdown
"Is this something you've always done, or did you discover it?"
"Does this feel like the real insight, or are we still on the surface?"

Phase 4: Session Export

When to Generate Export
  • User says "that's enough" or "let's wrap up"
  • Natural conclusion point reached
  • Sufficient breakthroughs documented (usually 2-4)
Output Format

You MUST return the session export in this exact format:

markdown
# Thought Partner Session

**Date:** [YYYY-MM-DD HH:mm]
**Topic:** [Brief description of what was explored]


## Narrative Arc

The journey through this session:

- **Starting Point:** [Where the conversation began]
- **First Turn:** [What shifted the direction]
- **Key Discovery #1:** [First breakthrough]
- **Deepening:** [How we went deeper]
- **Key Discovery #2:** [Second breakthrough]
- **Crystallization:** [How concepts got named]
- **Final Insight:** [Most powerful takeaway]


## Breakthroughs Summary

### Breakthrough 1: [Name of Concept]
[2-3 sentence summary of the insight]

**The paradox:** [If applicable]
**The pattern:** [What it reveals]
**Application:** [How to use this]

### Breakthrough 2: [Name of Concept]
[2-3 sentence summary]

[Continue for each breakthrough...]


## Named Concepts Dictionary

| Concept | Definition | Origin in Session |
|---------|------------|-------------------|
| [Name 1] | [Brief definition] | [Where it emerged] |
| [Name 2] | [Brief definition] | [Where it emerged] |


## Patterns Observed

- [Pattern 1]
- [Pattern 2]

## Paradoxes Discovered

- [Paradox 1]: [Conventional wisdom] vs [User's counterintuitive truth]
- [Paradox 2]: ...

## Potential Applications

- [How insight 1 could be applied to content, products, etc.]
- [How insight 2 could be applied]


## Session Transcript Highlights

### [Topic/Thread Headline]

**Partner:** [Key question or observation]

**User:** [Response that led somewhere]

**Partner:** [Follow-up that deepened]

**User:** [Breakthrough response]

[Continue with significant exchanges...]


## Next Steps

Based on this session, consider:
1. [Suggested next step]
2. [Another suggestion]
3. [Optional deeper exploration]

Redirect Handling

When user redirects, respond naturally:

User SaysPartner Response
"We're going in the wrong direction""Got it. What direction feels more right?"
"Switch topics""Sure. What else is on your mind?"
"I don't understand this""Let me try a different angle. [Rephrase]"
"This isn't landing""No problem. What would be more useful to explore?"
"I think we're done""Good session. Let me capture what we discovered."

Error Handling

User Provides No Clear Topic
markdown
"This could be a method you use, a belief you hold, something
you're building, or just an idea you've been turning over.

What's been on your mind lately that you'd like to explore?"
Conversation Goes Flat

Try a different technique:

  • If pattern spotting isn't working → try paradox hunting
  • If direct questions aren't working → make observations instead
markdown
"We might be on the surface still. What's something about
this that feels hard to explain to others?"
User Gets Stuck

Offer a bridge:

markdown
"Let me share what I'm noticing so far..."

[Summarize 2-3 patterns you've observed]

"Does any of that resonate? Or is there something else entirely?"
Insufficient Material for Breakthroughs
markdown
"We've covered good ground but haven't hit a breakthrough yet.

Options:
1. Go deeper on [specific area mentioned]
2. Try a different topic
3. Save what we have and continue another time

What feels right?"

Integration

With skill-content-pipeline

Use breakthroughs as source material for content creation.

With skill-meta-prompt

Generate prompts based on discovered concepts.

With skill-decision-support

When exploration reveals decision points, transition to options presentation.

With flow-discover

Research can feed into thought partner session for interpretation.

The Bottom Line

Thought partner → Observe → Question → Challenge → Name → Document
Otherwise → Surface-level conversation → Generic insights → Forgettable

Listen deeply. Question relentlessly. Name what you find.

© nyldn, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in skills/skill-thought-partner of nyldn/claude-octopus.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit e14b84f

Used in 1 other repository

We found 2 copies of this SKILL.md (exact, near-identical or edited) in other folders, from 1 other GitHub owner. This page covers the copy in nyldn/claude-octopus, which our catalogue first saw on October 7, 2026.

Compare with similar skills

Skill Thought Partner next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Skill Thought Partner compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Skill Thought Partner this skillnyldn/claude-octopus4.2k1 repos~3.3kAutomated safety check: PassMIT
Brainstormingxpinjection/test-driven-spring-boot11254 repos~2.6kAutomated safety check: PassMIT
Typesafe AIOpenAgentsInc/openagents4559 repos~2.5kAutomated safety check: PassMIT
Yao Meta Skillyaojingang/yao-meta-skill2.7k—~768Automated safety check: PassMIT
Trellis StartROYIANS/foliq-print-template-designer1356 repos~646Automated safety check: PassMIT
Brainstorming Before BuildingjnMetaCode/superpowers-zh8.3k—~1.8kAutomated safety check: PassMIT

Similar skills

  • Brainstorming

    xpinjection/test-driven-spring-boot

    You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior.

    112 GitHub starsUsed in 54 repos~2.6k tokens
    Agent WorkflowsAuto-check passed
  • Typesafe AI

    OpenAgentsInc/openagents

    Build AI-powered software with TypeSafe: small units of AI intelligence you can use like programming primitives.

    455 GitHub starsUsed in 9 repos~2.5k tokens
    Agent WorkflowsAuto-check passed
  • Yao Meta Skill

    yaojingang/yao-meta-skill

    Create, improve, or evaluate an existing skill from workflows, prompts, SOPs, scripts.

    2.7k GitHub stars~768 tokensUpdated 1 mo ago
    Agent WorkflowsAuto-check passed
  • Trellis Start

    ROYIANS/foliq-print-template-designer

    Initializes an AI development session by reading workflow guides, developer identity, git status, active tasks, and project guidelines from .trellis/.

    135 GitHub starsUsed in 6 repos~646 tokens
    Agent WorkflowsAuto-check passed
  • Brainstorming Before Building

    jnMetaCode/superpowers-zh

    Turns a rough idea into an approved design before any code is written, sorting the request into spike, bounded or architectural and enforcing an approval gate.

    8.3k GitHub stars~1.8k tokensUpdated yesterday
    Agent WorkflowsAuto-check passed
  • Om Brainstorm

    go-musicfox/go-musicfox

    Divergent conversation before any artifact exists — open questions one at a time, alternatives including building nothing, converging on a routing decision and a handoff brief for the next skill.

    2.6k GitHub starsUsed in 1 repo~1.6k tokens
    Agent WorkflowsAuto-check passed

More from nyldn/claude-octopus

All 62 skills in this repo
  • Octopus Quick

    nyldn/claude-octopus

    Quick execution for ad-hoc tasks without full workflow overhead — use for small, self-contained requests

    4.2k GitHub starsUsed in 1 repo~2.2k tokens
    Auto-check passed
  • Octopus Research

    nyldn/claude-octopus

    Thorough research across multiple sources — use for complex topics needing broad synthesis

    4.2k GitHub starsUsed in 1 repo~1.9k tokens
    Auto-check passed
  • Octopus Security Audit

    nyldn/claude-octopus

    OWASP compliance, vulnerability scanning, and adversarial red team testing — use for security reviews

    4.2k GitHub starsUsed in 1 repo~2.3k tokens
    Auto-check passed
  • Skill Audit

    nyldn/claude-octopus

    Audit codebases for quality, consistency, and broken patterns — use for pre-release or tech debt review

    4.2k GitHub starsUsed in 1 repo~3.2k tokens
    Auto-check passed
  • Skill Content Pipeline

    nyldn/claude-octopus

    Extract patterns and anatomy from URLs — use to reverse-engineer content strategies from live pages

    4.2k GitHub starsUsed in 1 repo~3.9k tokens
    Auto-check passed
  • Skill Context Detection

    nyldn/claude-octopus

    Auto-detect work context (Dev vs Knowledge) — use to tailor workflows based on current task type

    4.2k GitHub starsUsed in 1 repo~2.6k tokens
    Auto-check passed

Categories

Questions about Skill Thought Partner

What does Skill Thought Partner do?

Brainstorm creatively with pattern spotting and paradox hunting — use for ideation and exploration. Skill Thought Partner is an agent skill from nyldn/claude-octopus.

When should I use Skill Thought Partner?

Skill Thought Partner fits situations like: ideation and exploration; tasks that involve Brainstorming.

How do I install Skill Thought Partner in Claude Code?

Run `npx skills add nyldn/claude-octopus --skill skill-thought-partner -a claude-code`. Or copy the skill folder (skills/skill-thought-partner in nyldn/claude-octopus) into .claude/skills/skill-thought-partner in your project. Claude Code loads it when a task matches its description.

How do I install Skill Thought Partner in Codex?

Run `npx skills add nyldn/claude-octopus --skill skill-thought-partner -a codex`. Or copy the skill folder (skills/skill-thought-partner in nyldn/claude-octopus) into .agents/skills/skill-thought-partner in your project. Codex loads it when a task matches its description.

Can I use Skill Thought Partner in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add nyldn/claude-octopus --skill skill-thought-partner -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/skill-thought-partner, .gemini/skills/skill-thought-partner, .github/skills/skill-thought-partner and .opencode/skills/skill-thought-partner in your project.

What does Skill Thought Partner need to run?

SKILL.md names no scripts, command-line tools or credentials: Skill Thought Partner is instructions for the agent only.

Does Skill Thought Partner access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Skill Thought Partner safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Skill Thought Partner use?

Skill Thought Partner is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Skill Thought Partner use?

About 3.3k tokens (SKILL.md is roughly 13k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Skill Thought Partner?

Skills that share tags, products or a category with Skill Thought Partner: Brainstorming (xpinjection/test-driven-spring-boot, 112 stars), Typesafe AI (OpenAgentsInc/openagents, 455 stars), Yao Meta Skill (yaojingang/yao-meta-skill, 2.7k stars) and Trellis Start (ROYIANS/foliq-print-template-designer, 135 stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Skill Thought Partner?

nyldn (a GitHub user) maintains it in nyldn/claude-octopus, which has 4,192 GitHub stars. The repository holds 62 skills in this directory. The repository was last updated on October 9, 2026.

Source: nyldn/claude-octopus on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.