explore
Skill25 starsChanged 7 months agoMITView on GitHub
Search the marketplace, analyze competition, discover gaps, and find inspiration. Use when: 'explore', 'search marketplace', 'what exists for', 'competitors', 'market gap', 'trending', or before browsing the marketplace. For domain research with build recommendations, use /scout instead.
- Installs packages
69 other projects use a version of this.
Adopted by
Most recently taken up by
- Atmoskun/claude-skillstheir own version, first seen Nov 16, 2026
- timursafiullin/aodtheir own version, first seen Nov 15, 2026
- Renozoic-Foundry/forge-publictheir own version, first seen Nov 15, 2026
- lumpcode/lumpcodetheir own version, first seen Nov 14, 2026
- iwe-org/dev-workspacetheir own version, first seen Nov 14, 2026
- jskswamy/skillstheir own version, first seen Nov 8, 2026
Used with
Projects with it also use agent-context-isolation, agentica-infrastructure and 10 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
parcadeiSkillagent-context-isolation
Agent Context Isolation
Also in 3 projects that use this
parcadeiSkillagentica-infrastructure
Reference guide for Agentica multi-agent infrastructure APIs
Also in 3 projects that use this
- +10 more from parcadei
Other versions69 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 69
- parcadei/Continuous-Claude★ 3.9ksame name
- parcadei/Continuous-Claude-v2★ 3.9ksame name
- parcadei/Continuous-Claude-v3★ 3.9ksame name
- JordanCoin/codemap★ 715same name
- ZimoLiao/scholaraio★ 585same name
- strands-agents/box★ 547same name
- andreasmolnardev/dashwise-next★ 465same name
- ai-analyst-lab/ai-analyst★ 304same name
- exmergo/dex★ 132same name
- striderZA/OpenCodeGameStudios★ 124same name
59 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 6 are distinctive, and 3 are very specific to this project.
0
Common
Most of the 60 edited versions we read keep these.
6
Distinctive
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 21
When to use: Before /create (market research).
- Very specific: line 30
Read creator.yaml → adapt to profile.type, expertise domains, language
- Very specific: line 31
Check CLI: myclaude --version 2>/dev/null — if unavailable, explain install and offer offline alternatives
- Very specific: line 79
IMPORTANT: All detection uses Glob, Read, and Grep tools — NOT Bash.
- Distinctive: line 117
Anti-pattern guard: Don't recommend more than 5 products total.
- Distinctive: line 165
Run trending + search across all categories
- Distinctive: line 240
If myclaude CLI is not installed:
- Distinctive: line 276
When summarizing this conversation, always preserve:
- Distinctive: line 277
Search queries executed and key results
“Common” is measured: kept by at least 6 in 10 of the 60 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Explore — Marketplace Intelligence & Discovery
Before you build, know the landscape. This skill is your window into the marketplace.
When to use: Before /create (market research). After /publish (competitive positioning). Anytime you want to understand what exists, what's missing, and where the opportunity is.
When NOT to use: For building products (use /create→/fill pipeline). For quality checks (use /validate).
Activation Protocol
- Shared preamble: Load
references/quality/activation-preamble.md— context assembly, persona adaptation, deterministic routing rules. - Read
creator.yaml→ adapt to profile.type, expertise domains, language - Check CLI:
myclaude --version 2>/dev/null— if unavailable, explain install and offer offline alternatives - Parse
$ARGUMENTSfor query intent - Route to the appropriate exploration mode
- Load voice identity: Load
references/quality/engine-voice-core.md. Every user-facing line in this skill honors the ✦ signature, three tones, and six anti-patterns. 5b. Load UX vocabulary: Loadreferences/ux-vocabulary.md— translate all internal terms before creator-facing output. 5c. Load proactives: Loadreferences/engine-proactive.md— wire #1 (pipeline guidance: after explore, suggest /scout or /create), #14 (feedback loop on published products). - CLI contract: Load
references/cli-contract.mdfor unified error handling. This skill is the heaviest CLI consumer (12+ invocations across 5 modes). Severity map:- Silent-skip:
search,trending,workspace --recommend— skip without user-visible warning on failure - Silent-skip:
--versioncheck — degrade to offline mode - Interactive:
install {slug}— show manual install instructions on failure - All queries: append
--json 2>/dev/null, parse JSON, treat invalid JSON as CLI absent - Timeout: 15s per command. On timeout, skip with offline alternatives
- Silent-skip:
Exploration Modes
MODE 0: Project-Aware Recommendations (the smart default)
Trigger: /explore with no arguments, "what do I need", "recommend", "analyze my setup", "optimize"
This is the MODE that makes the Engine indispensable. Instead of generic search, it analyzes the creator's ACTUAL project and recommends products that would improve their specific workflow.
Step 1 — Detect active project context (using Glob/Read/Grep, NOT Bash):
# What tech stack? Use Glob to detect manifest files:
Glob("package.json") → Node.js/JavaScript
Glob("Cargo.toml") → Rust
Glob("requirements.txt") → Python (pip)
Glob("pyproject.toml") → Python (modern)
Glob("go.mod") → Go
Glob("Gemfile") → Ruby
Glob("pom.xml") → Java (Maven)
Glob("build.gradle") → Java/Kotlin (Gradle)
# What framework? If package.json found, Read it and check dependencies:
Read("package.json") → look for: react, next, vue, angular, svelte, express, fastapi in dependencies
# What's in their .claude/?
Glob(".claude/rules/*.md") → count rule files
Glob(".claude/skills/*/SKILL.md") → count installed skills
# Do they have hooks?
Read(".claude/settings.json") → check if "hooks" key exists
# CLAUDE.md size? Read first 5 lines to detect existence, check file metadata
Glob("CLAUDE.md") + Glob(".claude/CLAUDE.md") → if found, Read first 10 lines for size estimateIMPORTANT: All detection uses Glob, Read, and Grep tools — NOT Bash. The only Bash allowed is Bash(myclaude *) for marketplace queries.
Step 2 — Cross-reference with marketplace: Based on detected tech stack, search for matching products:
myclaude search "{detected_framework}" --json 2>/dev/null
myclaude search "{detected_language} rules" --json 2>/dev/nullStep 3 — Deliver personalized recommendations:
Project Analysis: {project_name or cwd}
Stack: {language} + {framework}
Rules: {N} loaded ({M} without path scoping)
Skills: {N} installed
Hooks: {present/none}
CLAUDE.md: {N} chars ({optimal/oversized})
Recommended products for THIS project:
INSTALL NOW (free, immediate value):
1. {name} — {why it matches this project}
myclaude install {slug}
CONSIDER (paid, high impact):
2. {name} — {why it matches}
myclaude install {slug}
MISSING FROM YOUR SETUP:
- No {type} for {detected_gap} — {N} options exist on marketplace
- Your CLAUDE.md {has no path scoping / is oversized / could use @include}
CREATE OPPORTUNITY:
- No {type} exists for {framework + specific use case}
- Your expertise in {domain} + this project's stack = potential productAnti-pattern guard: Don't recommend more than 5 products total. Don't recommend paid products to beginners. Don't recommend products the creator already has installed. If CLI unavailable, show only the project analysis part (still valuable without marketplace).
Why this is MODE 0 (default): When a creator types /explore with no arguments, THIS is what runs. Not a generic "what do you want to search?" — a smart, project-aware consultation that gives immediate value. The creator didn't come to browse. They came because they want their setup to be better. Give them that.
MODE 1: Search (find specific products)
Trigger: "search for", "find", "is there a", specific query terms
myclaude search "{query}" --json 2>/dev/nullDisplay results as:
Found {N} products for "{query}":
{name} [{type}] by @{author} — {downloads} dl — ${price}
"{description truncated to 80 chars}"
...
{If 0 results: "No products found. This could be an opportunity — or the query needs refining."}MODE 2: Trending (market pulse)
Trigger: "trending", "popular", "what's hot", "top products"
myclaude trending --json 2>/dev/nullDisplay with analysis:
Top 10 trending products:
# NAME TYPE AUTHOR DL PRICE
1. {name} {type} @{auth} {N} {price}
...
Patterns I notice:
- Most popular type: {type with most entries}
- Price sweet spot: {most common price point}
- Gap: {type or category NOT represented in top 10}MODE 3: Gap Analysis (find opportunities)
Trigger: "gaps", "opportunities", "what's missing", "what should I build"
- Run trending + search across all categories
- Cross-reference with creator.yaml expertise domains
- Identify underserved categories:
myclaude search --category skills --sort newest --limit 10 --json 2>/dev/null
myclaude search --category squads --sort newest --limit 10 --json 2>/dev/null
myclaude search --category systems --sort newest --limit 10 --json 2>/dev/nullMarket Gap Analysis for @{username}:
YOUR EXPERTISE: {domains from creator.yaml}
UNDERSERVED AREAS:
- {category}: only {N} products, {avg_downloads} avg downloads
→ You could build: {suggestion matching their expertise}
OVERSATURATED:
- {category}: {N} products, hard to differentiate
→ Only enter if you have genuine domain expertise
RECOMMENDED FIRST PRODUCT:
{type}: {suggestion} — because {reasoning linking expertise to gap}MODE 4: Competitive Intel (analyze specific niche)
Trigger: "competitors for", "compare with", "who else builds", "analyze category"
myclaude search --category {category} --sort downloads --limit 10 --json 2>/dev/nullDeep analysis:
Competitive Landscape: {category}
LEADERS (>10 downloads):
{name} by @{author} — {downloads} dl — {description}
Strengths: {inferred from description}
Weakness: {what's missing based on DNA patterns}
DIFFERENTIATION OPPORTUNITIES:
1. {gap none of the leaders fill}
2. {feature that would make your product stand out}
3. {pricing opportunity if all are free or all are paid}MODE 5: Inspiration (explore what's possible)
Trigger: "inspire me", "show me examples", "what can I build"
Combine workspace scan + marketplace data:
myclaude workspace --recommend --json 2>/dev/nullBased on your setup and expertise:
QUICK WINS (build in <1 hour):
- {type}: {suggestion} — similar products average {N} downloads
AMBITIOUS PROJECTS (build in 1 day):
- {type}: {suggestion} — no competition in this space
DREAM PRODUCT (build over a week):
- {type}: {suggestion} — highest potential impactOffline Mode (CLI unavailable)
If myclaude CLI is not installed:
Marketplace search requires the MyClaude CLI.
Install: npm i -g @myclaude-cli/cli
In the meantime, I can:
- Analyze your workspace for product ideas based on your skills and rules
- Review structural-dna.md to suggest high-impact product types
- Help brainstorm based on your creator.yaml expertise domains
What would be most useful?Anti-Patterns
- Research as procrastination — If the creator has been exploring for 3+ rounds without starting a product, nudge: "You've done great research. The market validates your idea. Ready to /create?"
- Copy, not compete — Never suggest copying a top product. Always frame as "differentiate by adding what they lack."
- Stale data assumption — Marketplace data is live. Don't cache or assume results from previous searches.
- Overwhelm — Show max 10 results. More isn't helpful — it's paralyzing.
Quality Gate
- Results are actionable (end with "here's what to do next")
- Adapted to creator's language and technical level
- Gap analysis connects to creator's expertise (not generic)
- CLI failures degrade gracefully to offline alternatives
Compact Instructions
When summarizing this conversation, always preserve:
- Search queries executed and key results
- Market gaps identified and their connection to creator expertise
- Competitive analysis findings (leaders, differentiators)
- The recommended product opportunity and reasoning
- Any decisions made based on exploration
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
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.

