--- name: intel description: Market intelligence in seconds. Research gaps, competitors, sentiment, pricing, trends, and pain points for any market or product. For builders who need business decisions, not just information. argument-hint: "[mode] [topic]" — modes: gaps, competitors, sentiment, pricing, trends, pain, full context: fork agent: Explore disable-model-invocation: true allowed-tools: WebSearch, Read, Write, AskUserQuestion --- # /intel — Market Intelligence in Seconds Research any market, competitor, or product. Get actionable business intelligence with evidence. ## Quick Start Examples ``` /intel gaps AI video editing /intel competitors Notion /intel sentiment Claude Code /intel pricing email marketing SaaS /intel trends no-code 2026 /intel pain project management /intel full personal finance apps ``` --- ## Step 1: Parse Input Parse the user's input: `/intel [MODE] [TOPIC]` **MODES:** | Mode | Purpose | |------|---------| | `gaps` | Find market gaps and unmet needs | | `competitors` | Map competitive landscape | | `sentiment` | Quantify love/hate ratio | | `pricing` | Research pricing strategies | | `trends` | Analyze direction and velocity | | `pain` | Find top frustrations | | `validate` | Quick idea validation (gaps + competitors + pain combined) | | `full` | Complete report (all modes) | **If no mode specified**, default to `full`. **Store:** - `MODE = [extracted mode]` - `TOPIC = [extracted topic]` --- ## Step 2: Execute Searches Run **5-8 searches per mode**. Keep queries SIMPLE — complex boolean often fails. ### Search Templates by Mode **GAPS MODE** — Find unmet needs: ``` Search 1: {TOPIC} "I wish" site:reddit.com Search 2: {TOPIC} frustrating OR annoying site:reddit.com Search 3: {TOPIC} "would pay for" site:reddit.com Search 4: {TOPIC} feature request site:news.ycombinator.com Search 5: {TOPIC} problems complaints 2025 2026 Search 6: {TOPIC} missing features gaps ``` **COMPETITORS MODE** — Map landscape: ``` Search 1: {TOPIC} vs alternative site:reddit.com Search 2: {TOPIC} competitors comparison 2026 Search 3: best {TOPIC} alternatives Search 4: {TOPIC} review site:producthunt.com Search 5: {TOPIC} pricing plans comparison Search 6: "switched from" {TOPIC} to site:reddit.com ``` **SENTIMENT MODE** — Quantify opinions: ``` Search 1: {TOPIC} love OR amazing site:reddit.com Search 2: {TOPIC} hate OR terrible site:reddit.com Search 3: {TOPIC} honest review 2025 2026 Search 4: {TOPIC} worth it site:reddit.com Search 5: {TOPIC} overrated OR underrated Search 6: {TOPIC} complaints site:twitter.com ``` **PRICING MODE** — Research pricing: ``` Search 1: {TOPIC} pricing cost 2026 Search 2: {TOPIC} free tier freemium Search 3: {TOPIC} expensive OR cheap site:reddit.com Search 4: {TOPIC} pricing comparison Search 5: {TOPIC} "how much does" cost Search 6: {TOPIC} pricing strategy SaaS ``` **TRENDS MODE** — Analyze direction: ``` Search 1: {TOPIC} trending growing 2026 Search 2: {TOPIC} declining dying site:reddit.com Search 3: {TOPIC} future prediction 2026 Search 4: {TOPIC} news announcement 2026 Search 5: {TOPIC} hype OR bubble Search 6: {TOPIC} market size growth ``` **PAIN MODE** — Find frustrations: ``` Search 1: {TOPIC} frustrating painful site:reddit.com Search 2: {TOPIC} biggest problem site:reddit.com Search 3: {TOPIC} workaround hack site:reddit.com Search 4: {TOPIC} rant vent site:reddit.com Search 5: {TOPIC} broken buggy unreliable Search 6: {TOPIC} complaints issues 2026 ``` **VALIDATE MODE** — Quick idea validation (run in sequence): ``` # Pain check Search 1: {TOPIC} frustrating problem site:reddit.com Search 2: {TOPIC} "I wish" OR "would pay for" site:reddit.com # Competition check Search 3: {TOPIC} vs alternatives comparison Search 4: best {TOPIC} tools 2026 # Gap check Search 5: {TOPIC} missing features gaps complaints Search 6: {TOPIC} "doesn't exist" OR "no good" site:reddit.com ``` **FULL MODE** — Run 2-3 searches from EACH mode above. --- ## Step 3: Extract & Synthesize For each result, extract: - **Source URL** (required) - **Platform** (Reddit, HN, Twitter, blog) - **Key quote** (most relevant 1-2 sentences) - **Engagement hint** (if URL contains /comments/ or shows votes in snippet) ### Weighting Rules 1. **Reddit/HN** → Highest weight (real discussions) 2. **Twitter/X** → High (real-time sentiment) 3. **Product Hunt** → Medium (launch feedback) 4. **Blogs** → Medium (detailed but may be biased) 5. **Company sites** → Lowest (facts only, not opinions) ### Pattern Detection - What appears in **3+ sources**? → Strong signal - What has **engagement hints** (comments, upvotes mentioned)? → Validated - **Contradictions** between sources? → Note them - **Data gaps**? → Acknowledge low confidence --- ## Step 4: Output Format **CRITICAL: Follow this exact format.** ```markdown ## /intel {MODE}: {TOPIC} ### TL;DR [One paragraph — THE key insight a founder needs. Be specific. Be actionable.] ### Key Findings **1. [Finding title]** [2-3 sentences with SPECIFIC evidence] > "[Direct quote from source]" — [Source: platform/url] **2. [Finding title]** [2-3 sentences with evidence] > "[Direct quote]" — [Source] **3. [Finding title]** [Evidence and quotes] [3-5 findings total] ### Data Summary | Metric | Value | Notes | |--------|-------|-------| | Sources analyzed | {n} | Reddit: {n}, HN: {n}, Other: {n} | | Strong signals | {n} | Appeared in 3+ sources | | Confidence | {HIGH/MED/LOW} | Based on source quality | {MODE-SPECIFIC SECTION — see below} ### Strategic Implications **If building in this space:** - [Specific advice for founders/builders] **If investing/evaluating:** - [Specific advice for investors/analysts] **Contrarian take:** - [What the data might be missing or getting wrong] ### Confidence Breakdown | Factor | Score | Reasoning | |--------|-------|-----------| | Source diversity | {1-10} | {n} independent sources | | Data recency | {1-10} | {%} from last 90 days | | Signal strength | {1-10} | {n} findings in 3+ sources | | **Overall** | **{1-10}** | | ### Limitations & Blind Spots - [Data gaps — what's NOT represented] - [Bias warnings — who's overrepresented] - [Recency issues — old data flagged] --- 📊 /intel {MODE} complete ├─ Topic: {TOPIC} ├─ Sources: {n} total └─ Confidence: {HIGH/MED/LOW} Want deeper analysis? Ask about any finding. ``` --- ## Mode-Specific Sections ### GAPS Mode Output ```markdown ### Market Opportunities | Rank | Gap | Signal Strength | Evidence | |------|-----|-----------------|----------| | 1 | [Gap description] | {n} mentions | "[quote]" | | 2 | [Gap] | {n} mentions | "[quote]" | | 3 | [Gap] | {n} mentions | "[quote]" | ### Recommended Action [1-2 sentences: Which gap to pursue and why] ``` ### COMPETITORS Mode Output ```markdown ### Competitive Landscape | Competitor | Positioning | Mentioned | Sentiment | |------------|-------------|-----------|-----------| | [Name] | [1-line] | {n}x | 👍/👎/😐 | | [Name] | [1-line] | {n}x | 👍/👎/😐 | ### Positioning Gap [Where is there room in the market?] ``` ### SENTIMENT Mode Output ```markdown ### Sentiment Analysis **Score: {1-10}/10** (estimated from {n} sources) | Aspect | Positive | Negative | |--------|----------|----------| | [Aspect 1] | {n} mentions | {n} mentions | | [Aspect 2] | {n} | {n} | **Most Loved:** [Feature] — "[quote]" **Most Hated:** [Feature] — "[quote]" **Trend:** {Improving/Stable/Declining} ``` ### PRICING Mode Output ```markdown ### Pricing Landscape | Tier | Range | Model | Examples | |------|-------|-------|----------| | Entry | $X-Y | [model] | [products] | | Mid | $X-Y | [model] | [products] | | Premium | $X-Y | [model] | [products] | **Premium Justifiers:** [What features command higher prices] **Common Mistakes:** [Pricing errors to avoid] ``` ### TRENDS Mode Output ```markdown ### Trend Analysis **Direction:** {📈 Growing | 📊 Stable | 📉 Declining} **Velocity:** {Rapid | Moderate | Slow} | Signal | Evidence | |--------|----------| | [Trend signal] | "[quote]" | | [Signal] | "[quote]" | **Forecast:** [1-2 sentences on where this is heading] ``` ### PAIN Mode Output ```markdown ### Pain Point Rankings | Rank | Pain Point | Frequency | Severity | |------|------------|-----------|----------| | 1 | [Pain] | {n} mentions | High/Med/Low | | 2 | [Pain] | {n} mentions | High/Med/Low | | 3 | [Pain] | {n} mentions | High/Med/Low | ``` ### VALIDATE Mode Output Quick idea validation — combines gaps, competitors, and pain into a decision framework. ```markdown ### Idea Validation: {TOPIC} #### 1. Is There a Real Problem? **Verdict:** {YES/MAYBE/NO} - Pain points found: {n} - Top pain: [description] — "[quote]" - Severity: {High/Med/Low} #### 2. Are There Existing Solutions? **Verdict:** {CROWDED/SOME/EMPTY} - Competitors found: {n} - Top players: [list] - User satisfaction: {High/Med/Low} #### 3. Is There Room for You? **Verdict:** {YES/MAYBE/NO} - Gaps identified: {n} - Biggest gap: [description] - Differentiation opportunity: [specific angle] ### Overall Validation Score | Criteria | Score | Notes | |----------|-------|-------| | Problem exists | {1-10} | | | Market not saturated | {1-10} | | | Clear differentiation | {1-10} | | | **BUILD/DONT BUILD** | **{verdict}** | | ### Recommendation [2-3 sentences: Should they build this? Why or why not? What angle?] **Current Workarounds:** - [Pain 1]: People do [workaround] **Biggest Opportunity:** [Which pain to solve first] ``` ### FULL Mode Output Include ALL sections in order: 1. TL;DR (synthesis) 2. Market Gaps 3. Competitive Landscape 4. Sentiment Overview 5. Pricing Intel 6. Trend Direction 7. Pain Points 8. Overall Recommendation --- ## Error Handling **If search returns few/no results:** ```markdown ## /intel {MODE}: {TOPIC} ### Limited Data Available Only {n} relevant sources found. This could mean: - Niche/emerging topic with sparse coverage - Try broader search terms - Topic may not have significant online discussion **What I found:** [List whatever was found] **Suggestions:** - Try: /intel {MODE} [broader term] - Try: /intel {MODE} [related term] ``` **If topic is ambiguous:** Ask clarifying question before searching: ``` The topic "{TOPIC}" could mean several things: 1. [Interpretation 1] 2. [Interpretation 2] Which should I research? ``` --- ## Quality Checklist Before outputting, verify: - [ ] Every finding has a source URL - [ ] Numbers are specific ("12 mentions" not "many") - [ ] Direct quotes are actual quotes from sources - [ ] Confidence level matches data quality - [ ] Format matches templates exactly - [ ] Actionable — founder can decide something --- ## Context Memory After research, remember: - `MODE`: {mode} - `TOPIC`: {topic} - `KEY_FINDINGS`: Top 3-5 insights - `SOURCES`: URLs you found **You are now an expert on this topic.** For follow-ups: - Answer from your research (don't re-search) - Cite the sources you found - Offer to dive deeper on any finding Only new research if user asks about DIFFERENT topic. --- ## Anti-Patterns to Avoid ❌ **Don't hallucinate sources** — If you didn't find it, don't cite it ❌ **Don't use vague numbers** — "Several" "Many" "Some" → Use actual counts ❌ **Don't skip the format** — Follow templates exactly ❌ **Don't over-promise** — Low data = say "low confidence" ❌ **Don't re-search on follow-ups** — Use what you already found --- ## Reference Examples See the `examples/` folder for complete output examples: - `examples/gaps-example.md` — AI video editing gaps analysis - `examples/competitors-example.md` — Notion competitive landscape - `examples/sentiment-example.md` — Claude Code sentiment analysis Use these as templates for consistent output formatting. --- **/intel** — Stop guessing. Start knowing.