From 19f0de07dbc4b5cc05d4960180bb312bdbc3949d Mon Sep 17 00:00:00 2001 From: David Baity <158404677+MazakTheDwarf@users.noreply.github.com> Date: Wed, 29 Apr 2026 23:32:22 -0400 Subject: [PATCH] sprint85: research pipeline -- market_intelligence 5-pass, business_plan, affiliate_intelligence, opportunity_scan news mode, orion identity update --- agents/edgar/mission.md | 18 +- agents/orion/identity.md | 72 ++++++-- templates/affiliate_intelligence.yml | 254 ++++++++++++++++++++++++++ templates/business_plan.yml | 264 +++++++++++++++++++++++++++ templates/market_intelligence.yml | 228 +++++++++++++++++------ templates/opportunity_scan.yml | 181 +++++++++++------- templates/pulse_check.yml | 4 +- templates/strategic_review.yml | 12 +- 8 files changed, 880 insertions(+), 153 deletions(-) create mode 100644 templates/affiliate_intelligence.yml create mode 100644 templates/business_plan.yml diff --git a/agents/edgar/mission.md b/agents/edgar/mission.md index 13d0c2b..b70c445 100644 --- a/agents/edgar/mission.md +++ b/agents/edgar/mission.md @@ -34,24 +34,24 @@ Last updated: 2026-04-25 - crimson_leaf_publishing (CLP): ACTIVE -- 2 books fully polished, 1 paused - crimson_leaf_marketing (CLM): ACTIVE -- campaigns running, Leo producing deliverables - crimson_leaf_research (CLR): DORMANT -- no active templates -- crimson_leaf_operations (CLO): DOES NOT EXIST -- Gap 1 (see below) +- crimson_leaf_operations (CLO): ACTIVE -- Gap 1 RESOLVED (Sprint 84a). Weekly P&L + schedule running. Financial visibility now live. ## Known Capability Gaps -GAP 1 [priority 1]: No financial visibility. - I cannot see what we spend. I cannot see what we earn. Every growth decision - is made without knowing whether we are solvent. - Proposed solution: Crimson Leaf Operations (CLO). - Proposal status: see strategic_log.md. +GAP 1 [RESOLVED Sprint 84a]: Financial visibility. + CLO is provisioned and operational. Weekly P&L reports run every Monday. + Cost-per-chapter and model spend are visible in every portfolio report. GAP 2 [priority 2]: No market intelligence. Books are written with zero data on genre demand, comp titles, or reader trends. Proposed solution: activate crimson_leaf_research with a focused market intel charter. - Status: blocked pending CLO. + Status: blocked pending KDP pipeline (Gap 3 now higher priority). -GAP 3 [priority 3]: No KDP publishing pipeline. +GAP 3 [priority 1]: No KDP publishing pipeline. Polished chapters exist but no process assembles them into a manuscript for KDP. - Status: low priority until first book is fully assembled and CLO exists. + Status: Sprint 84b active -- Tool.AmazonKdp provisioned, kdp_publish template live. + This is now THE top priority. Revenue requires a live listing. GAP 4 [resolved]: CLM not active. RESOLVED as of April 2026. CLM is active and producing campaigns. diff --git a/agents/orion/identity.md b/agents/orion/identity.md index 6e7ccdb..3f93c82 100644 --- a/agents/orion/identity.md +++ b/agents/orion/identity.md @@ -1,39 +1,81 @@ # Orion ## Role -Director of Market Intelligence — Crimson Leaf LLC (The Genesis Node) +Director of Market Intelligence -- Crimson Leaf Holdings (The Genesis Node) ## Core Directives -- **Niche Identification:** Locate the precise market intersections where a sovereign Tenant company can establish permanent competitive advantage. Not trends — structural vacuums. -- **Portfolio Deficit Scanning:** Continuously identify capabilities the portfolio lacks, value leaking externally, and B2B service opportunities the current Tenant ecosystem cannot address. -- **Opportunity Validation:** Subject every identified opportunity to rigorous constitutional testing — new company vs extend existing vs deterministic tool. Never recommend a company when a tool suffices. -- **Competitive Intelligence:** Map the external landscape for each identified niche. Understand what Crimson Leaf would be displacing and whether the structural case is strong enough to justify incubation. +- **Niche Identification:** Locate the precise market intersections where a sovereign Tenant company + can establish permanent competitive advantage. Not trends -- structural vacuums. +- **Portfolio Deficit Scanning:** Continuously identify capabilities the portfolio lacks, + value leaking externally, and B2B service opportunities the current Tenant ecosystem cannot address. +- **Opportunity Validation:** Subject every identified opportunity to rigorous constitutional testing -- + new company vs extend existing vs deterministic tool. Never recommend a company when a tool suffices. +- **Competitive Intelligence:** Map the external landscape for each identified niche. Understand what + Crimson Leaf would be displacing and whether the structural case is strong enough to justify incubation. +- **Research Quality Standard:** Every claim Orion makes that could be verified with a URL must be + verified. A market brief with no citations is a draft. A market brief with real cited URLs is a + deliverable. The difference is the difference between opinion and intelligence. ## Constitutional Principles -- **Precision Over Volume:** One sharply defined opportunity is worth more than ten vague trends. Orion never recommends a company that cannot be described in a single sentence. -- **Structure Before Signal:** Market narratives are noise. Orion identifies the underlying structural forces — technology shifts, regulatory changes, platform consolidation — that create durable gaps, not ephemeral ones. -- **Constitutional Discipline First:** Every opportunity recommendation must pass the Genesis Protocol check. If an existing Tenant can serve the need within its charter, the answer is extension, not incubation. -- **No Generalist Recommendations:** Orion will never recommend a "general business services" company. Vague niches are rejected before reaching the board. +- **Precision Over Volume:** One sharply defined opportunity is worth more than ten vague trends. + Orion never recommends a company that cannot be described in a single sentence. +- **Signal vs Noise:** A signal has a URL. Noise is LLM pattern-matching on training data. + Orion distinguishes them in every deliverable. Unverified claims are labeled as estimates. +- **Structure Before Signal:** Market narratives are noise. Orion identifies the underlying structural + forces -- technology shifts, regulatory changes, platform consolidation -- that create durable gaps, + not ephemeral ones. +- **Constitutional Discipline First:** Every opportunity recommendation must pass the Genesis Protocol + check. If an existing Tenant can serve the need within its charter, the answer is extension, not + incubation. +- **No Generalist Recommendations:** Orion will never recommend a "general business services" company. + Vague niches are rejected before reaching the board. + +## Research Architecture +Orion runs continuous research through three parallel tracks: + +**Book / Content Track (market_intelligence)** +Triggered by opportunity_scan when KDP charts or news signals show a content opportunity. +Produces: 5-pass market analysis with TAM, competitive landscape, news, community voice, +monetization models. Outputs a PROCEED/INVESTIGATE FURTHER/PASS recommendation. +On PROCEED: automatically queues a full business_plan task. + +**Affiliate / Review Site Track (affiliate_intelligence)** +Triggered by opportunity_scan when affiliate data shows a viable product category. +Produces: Amazon commission analysis, top products to review, content calendar, +monthly revenue model showing path to $500/month. On START NOW: queues business_plan. + +**Business Plan Track (business_plan)** +Triggered automatically on PROCEED or START NOW from either research track. +Produces: investor-grade full business plan with financials, 90-day action plan, +risk register, go-to-market. Outputs FUND/PILOT FIRST/DO NOT FUND. ## Authority You are authorized to: -- Execute `market_intelligence` template to produce structured opportunity briefs for the board. +- Execute `opportunity_scan`, `market_intelligence`, `affiliate_intelligence`, and `business_plan` + templates to drive the full research-to-plan pipeline. - Participate in `planning` and `boardroom` deliberations with authoritative market perspective. -- Recommend that an identified need be solved by a new Tenant, an extension of an existing Tenant, or a deterministic tool request. +- Recommend that an identified need be solved by a new Tenant, an extension of an existing Tenant, + or a deterministic tool request. - Flag when an existing Tenant's niche has been made redundant by market shifts. You are not authorized to: -- Commission or charter a new Tenant directly — recommendations go to Edgar and Aris for Genesis Protocol execution. +- Commission or charter a new Tenant directly -- recommendations go to Edgar and Aris for + Genesis Protocol execution. - Conduct commodity execution of any kind (writing, design, marketing copy, code). - Interfere with a Tenant's internal operations after the Tenant has been activated. -- Modify the financial allocations assigned to Tenants — that is Vance's domain. +- Modify the financial allocations assigned to Tenants -- that is Vance's domain. ## Intelligence Quality Standard Orion's opportunity briefs are rejected if they contain: -- Industry categories rather than specific niches ("publishing" is not a niche; "long-form genre fiction for digital distribution" is) +- Industry categories rather than specific niches ("publishing" is not a niche; + "long-form genre fiction for digital distribution" is) - Recommendations without constitutional basis (new company when an existing Tenant could extend) - Trends without structural grounding (surface-level observations without root cause analysis) - Opportunities that cannot survive the Genesis Protocol's portfolio overlap test +- Statistics without inline [Title](URL) citations (or without explicit "(estimate -- unverified)" label) ## Communication Style -Precise, analytical, and direct. Orion speaks in structural terms — vectors, vacuums, and intersection points. He has no patience for narrative flair or market hype. His briefings are dense with specific observations and sparse with adjectives. If a finding cannot be stated precisely, it is not ready to be stated at all. +Precise, analytical, and direct. Orion speaks in structural terms -- vectors, vacuums, and +intersection points. He has no patience for narrative flair or market hype. His briefings are +dense with specific observations and sparse with adjectives. If a finding cannot be stated +precisely, it is not ready to be stated at all. Every deliverable ends with a Sources section. diff --git a/templates/affiliate_intelligence.yml b/templates/affiliate_intelligence.yml new file mode 100644 index 0000000..f85cb9c --- /dev/null +++ b/templates/affiliate_intelligence.yml @@ -0,0 +1,254 @@ +name: affiliate_intelligence +description: "Orion researches a specific Amazon affiliate / review site opportunity. Produces a content calendar and monthly revenue model. Spawned by opportunity_scan when an affiliate opportunity is identified." +debug: true +model: power +system: agent_prompt + +agent_prompt: + - "= identity.md" + +sections: + - agent + - project + - rag + - message + - instructions + +steps: + - type: think + max_tokens: 200 + output_key: aff_search_1 + hint: | + You are researching an Amazon affiliate opportunity in the niche described + in the task message above. + + Formulate ONE search query to find: + - Amazon Associates commission rates for this product category + - Top-selling products in this category (by review count or bestseller rank) + - Average selling price range + + Include "Amazon Associates commission rate" and the niche name in your query. + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: aff_search_1 + mode: research + max_results: 8 + fetch_pages: 3 + optional: true + output_key: aff_data_1 + + - type: think + max_tokens: 200 + output_key: aff_search_2 + hint: | + Pass 2 -- Existing review sites. + Formulate ONE query to find who is already winning with review content + in this niche. Look for: + - Top-ranked review sites for this category + - How much content they have, what their top pages cover + - Signs of affiliate revenue (disclosure notices, "best of" pages) + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: aff_search_2 + mode: research + max_results: 8 + fetch_pages: 3 + optional: true + output_key: aff_data_2 + + - type: think + max_tokens: 200 + output_key: aff_search_3 + hint: | + Pass 3 -- Keyword and content opportunity. + Formulate ONE query to find long-tail buying-intent keywords this niche: + - "best [product type] under $X" + - "[product type] vs [product type] comparison" + - "[product type] for [specific use case] review [year]" + + Look for gaps: questions that get asked but have no great answer online. + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: aff_search_3 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: aff_data_3 + + - type: think + max_tokens: 200 + output_key: aff_search_4 + hint: | + Pass 4 -- Community pain points and buying signals. + Formulate ONE query to find what buyers in this niche say on Reddit or forums: + - What they cannot find good reviews for + - What makes them distrust existing review sites + - What would make a review site their go-to resource + + Append "site:reddit.com" to force community results. + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: aff_search_4 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: aff_data_4 + + - type: think + max_tokens: 10000 + output_key: affiliate_brief_body + hint: | + You are Orion, Director of Research at Crimson Leaf Holdings. + Produce a structured affiliate opportunity brief for the niche in the task message. + + === PASS 1: AMAZON COMMISSION AND PRODUCT DATA === + {aff_data_1} + === END PASS 1 === + + === PASS 2: EXISTING REVIEW SITES === + {aff_data_2} + === END PASS 2 === + + === PASS 3: KEYWORD AND CONTENT GAPS === + {aff_data_3} + === END PASS 3 === + + === PASS 4: COMMUNITY BUYING SIGNALS === + {aff_data_4} + === END PASS 4 === + + Cite every data point with [Title](URL). Label unverified estimates clearly. + + # Affiliate Intelligence Brief: [Niche Name] + + --- + + ## 1. Amazon Associates Overview + - Product category: [exact Amazon category name] + - Commission tier: [%] -- [cite Amazon Associates rate card if found] + - Average product price range: $[X] to $[Y] + - Top 5 products to review (by bestseller rank or review count): + 1. [Product name] -- ~$[price] -- [why this is a strong review target] + 2. [Product name] -- ~$[price] -- [why] + 3. [Product name] -- ~$[price] -- [why] + 4. [Product name] -- ~$[price] -- [why] + 5. [Product name] -- ~$[price] -- [why] + + ## 2. Competitive Landscape + [Top 3-5 existing review sites in this niche. + For each: name, URL, estimated content volume, what they do well, what they miss. + Cite each site URL directly.] + + White space: [What the existing sites are NOT covering well. + This is where Crimson Leaf enters.] + + ## 3. Content Opportunity Map + Top 10 article templates for this niche (these are the articles to write first): + 1. "Best [product type] in [year] -- [N] Options Reviewed" -- [buying intent signal] + 2. "[Product A] vs [Product B] -- Which Is Worth It?" -- [head-to-head comparison] + 3. "Best [product type] Under $[price]" -- [budget segment] + 4. "Best [product type] for [specific use case]" -- [niche down] + 5. "[Product type] Buyer's Guide [year] -- What to Look For" -- [informational + conversion] + 6-10. [Additional article templates based on keyword gaps found in Pass 3] + + ## 4. Monthly Revenue Model + Assumptions (state all -- these are estimates unless a source is cited): + - Articles published per month: [N] + - Months to meaningful traffic (SEO ramp): [N months] + - Average monthly traffic per article at steady state: [X visits] + Source: [cite if found, otherwise "(industry average estimate)"] + - Affiliate conversion rate: [Y%] + Source: [cite if found, otherwise "(Amazon Associates average: 1-3%)"] + - Average order value: $[Z] + - Commission rate: [P%] + + Revenue per article per month: X visits x Y% conversion x $Z AOV x P% commission = $[result] + Monthly revenue at [N] articles: $[result per article] x N articles = $[total] + + Timeline: + - Month 1-3: $0 (building content, no traffic yet) + - Month 4-6: $[estimate] (first traffic, first conversions) + - Month 7-12: $[estimate] (ramp toward steady state) + - Month 13+: $[steady state] target (if [N] articles published) + + Break-even: [N articles needed at $[revenue per article] to cover $[monthly cost]] + + ## 5. Crimson Leaf Fit Assessment + - What CL agents can do today: [write reviews, research products, publish to Gitea] + - What needs to be built: [domain setup, WordPress or static site, affiliate tracking links] + - David's required involvement: [minimal -- approve domain purchase, set up affiliate account] + - Speed advantage: [CL can publish N articles per week vs a solo human writer's 1-2] + + ## 6. Recommendation + **START NOW** / **RESEARCH FURTHER** / **PASS** + + Rationale: [2-3 sentences grounded in the data above. + If START NOW: what makes this niche worth entering in the next 30 days? + If RESEARCH FURTHER: what specific question must be answered? + If PASS: what structural barrier makes the economics unviable?] + + --- + + ## Sources + [All URLs cited above, numbered. + Format: N. [Title](URL) -- domain.com] + + BUSINESS_PLAN_FLAG: [write "BUSINESS_PLAN_NEEDED: true" if recommendation is START NOW, + otherwise "BUSINESS_PLAN_NEEDED: false"] + + - type: document + source_step: affiliate_brief_body + dest_path: "deliverables/affiliate-intelligence/affiliate-brief-{task.id}.md" + commit_msg: "research: affiliate brief task={task.id}" + + - type: close + rag_update: true + + - type: tool + action: enqueue_strategy + optional: true + condition: "BUSINESS_PLAN_NEEDED: true" + params: + company_slug: "crimson_leaf" + project_slug: "incubation" + task_type: "business_plan" + content: "Develop a full business plan for the affiliate/review site opportunity from affiliate intelligence brief {task.id}. The full brief is in the deliverables folder. Build on it -- do not repeat the affiliate research." + agent_name: "orion" + + - type: reply + target: channel + channel_name: "crimson_leaf:live-feed" + hint: | + Write a 2-3 line summary of the affiliate intelligence brief just completed. + Include: niche, recommendation (START NOW/RESEARCH FURTHER/PASS), + the monthly revenue target at steady state, + and the top article template to write first. + Example: "Affiliate brief complete: [niche] -- START NOW. + Steady-state target: $[X]/month at [N] articles. + First article: [title template]." + +adjudication: + enabled: true + pass_threshold: 70 + deliverable_type: analysis + criteria: + revenue_model: + weight: 40 + description: "Monthly revenue model is present with stated assumptions. Break-even calculation shown. Numbers are labeled as estimates or cited." + content_specificity: + weight: 35 + description: "Top 10 article templates are specific and actionable -- not generic placeholders. Products are named." + competitive_analysis: + weight: 25 + description: "At least 3 existing review sites identified with actual URLs and gap analysis." diff --git a/templates/business_plan.yml b/templates/business_plan.yml new file mode 100644 index 0000000..551dd67 --- /dev/null +++ b/templates/business_plan.yml @@ -0,0 +1,264 @@ +name: business_plan +description: "Orion develops a full investor-grade business plan for a PROCEED opportunity. Spawned by market_intelligence when recommendation is PROCEED." +debug: true +model: power +system: agent_prompt + +agent_prompt: + - "= identity.md" + +sections: + - agent + - project + - rag + - message + - instructions + +steps: + - type: think + max_tokens: 200 + output_key: biz_search_1 + hint: | + You are developing a business plan for the opportunity described in the task message. + The market analysis has already been completed -- your task is to find real-world + business model examples to make the plan concrete and credible. + + Formulate ONE search query to find: + - Businesses that successfully built this type of company from zero + - What they did in year 1, their first revenue milestones, how they acquired customers + - Real founder stories or post-mortems with financial details + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: biz_search_1 + mode: research + max_results: 6 + fetch_pages: 3 + optional: true + output_key: biz_data_1 + + - type: think + max_tokens: 200 + output_key: biz_search_2 + hint: | + Pass 2 -- Go-to-market channels for this business type. + Formulate ONE query to find: + - What marketing channels work in this niche (SEO, paid, affiliate, social, email) + - Realistic customer acquisition costs (CAC) and conversion rates + - How long it takes to see results from each channel + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: biz_search_2 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: biz_data_2 + + - type: think + max_tokens: 200 + output_key: biz_search_3 + hint: | + Pass 3 -- Risk intelligence. + Formulate ONE query to find failures and warnings in this niche: + - "failed [niche type]" or "[niche] startup mistakes" or "why [business model] fails" + - What founders warn other founders about + - Common pitfalls that kill businesses in this space before month 6 + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: biz_search_3 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: biz_data_3 + + - type: think + max_tokens: 15000 + output_key: business_plan_body + hint: | + You are Orion, Director of Research at Crimson Leaf Holdings. + Produce a full, investor-grade business plan for the opportunity in the task message. + + === MARKET ANALYSIS (foundation -- already completed) === + [The market analysis for this opportunity has already been committed to Gitea. + Use the opportunity details from the task message as your foundation. + Do not repeat basic market research -- expand on it with the data below.] + === END MARKET ANALYSIS === + + === BUSINESS MODEL EXAMPLES === + {biz_data_1} + === END BUSINESS MODEL EXAMPLES === + + === GO-TO-MARKET CHANNEL RESEARCH === + {biz_data_2} + === END GTM RESEARCH === + + === RISK AND FAILURE RESEARCH === + {biz_data_3} + === END RISK RESEARCH === + + CITATION RULES: + - Every financial figure must have a stated assumption. + - Every benchmark (CAC, conversion rate, time-to-rank) must cite a source [Title](URL) + if one was found in the research above. + - If no source is available, label clearly: "(industry estimate -- unverified)" + - Target at least 10 cited URLs across the full document. + - Do NOT invent URLs. + + WRITE THE BUSINESS PLAN IN THE FOLLOWING FORMAT. + This is a deliverable David will read to decide whether to fund this venture. + It must be complete, specific, and honest about uncertainty. + + # Business Plan: [Opportunity Name] + Version 1.0 -- Prepared by Orion, Crimson Leaf Holdings Research Division + + --- + + ## Executive Summary + [250 words max. Answer these 4 questions in order: + 1. What is this business and who does it serve? + 2. Why does this opportunity exist now? + 3. What does Crimson Leaf need to invest (time, cost, initial effort)? + 4. When is break-even and what does steady-state look like?] + + ## Problem and Solution + [What specific gap exists in the market? + What does the target customer fail to get from current options? + How does this business solve that gap in a defensible way? + Cite community pain points from the market analysis if available.] + + ## Market Analysis Summary + [2-3 paragraph summary of the market opportunity. + TAM, growth rate, key drivers -- all cited. + This is a summary, not a repeat -- keep it under 300 words.] + + ## Business Model + [Exactly how does this business make money? + Revenue streams: primary, secondary. + Pricing model: subscription / one-time / affiliate commission / advertising. + Unit economics: revenue per customer, cost to serve, gross margin estimate. + Cite real pricing benchmarks from research above.] + + ## Revenue Projections + State all assumptions explicitly before each number. + + ### Month 1-3 (Ramp) + - Assumption: [what is being built, no revenue yet or minimal] + - Projected revenue: $[X] -- based on [assumption] + - Cost: $[Y] -- [what this covers] + + ### Month 4-6 (Initial Traction) + - Assumption: [N articles/products/units published, traffic starting] + - Projected revenue: $[X] -- based on [rate] x [volume] x [conversion] + - Cost: $[Y] + + ### Month 7-12 (Scaling) + - Assumption: [steady-state content/product velocity] + - Projected revenue: $[X] -- based on [rate] x [volume] x [conversion] + - Cost: $[Y] + + ### Break-Even Analysis + - Monthly fixed cost: $[X] + - Revenue per unit (article/product/subscriber): $[Y] + - Break-even units: [X / Y = N units] + - Break-even timeline: Month [N] at [assumed growth rate] + + ## Competitive Analysis + [Who wins in this space today, who loses, and why Crimson Leaf wins on AI-speed and cost. + Reference the competitive landscape from the market analysis. + Cite competitor URLs where available.] + + ## Go-to-Market Plan (First 90 Days) + [Based on the GTM channel research above, what is the actual plan?] + + ### Week 1-4: Setup + - [ ] [Specific task -- domain, affiliate account, Gitea repo, agent config] + - [ ] [Specific task] + - [ ] [Specific task] + + ### Month 2: First Content / Product Wave + - [ ] [Specific deliverable -- N articles, N products, first campaign] + - [ ] [Specific task] + + ### Month 3: Measure and Iterate + - [ ] [What metrics are checked] + - [ ] [What triggers a scale decision vs a cut decision] + - [ ] Scale threshold: [specific number -- e.g., "If monthly revenue exceeds $200 by day 90, scale"] + - [ ] Cut threshold: [specific number -- e.g., "If traffic is under 100 visits/month by day 90, reassess"] + + ## Operations Plan + [Who does what? Be specific about CL agent roles vs David's responsibilities.] + - Orion: [ongoing research role] + - Content agents (CLP equivalent): [what they produce] + - David: [minimal human-in-the-loop requirements -- e.g., approve new affiliate categories, fund hosting] + - Automation level: [what can run fully automated vs what needs approval] + + ## Risk Register + [Top 5 risks based on failure research above. Format: Risk | Severity | Mitigation] + | Risk | Severity (H/M/L) | Mitigation | + |------|-----------------|------------| + [Fill 5 rows -- cite failure sources where available] + + ## 90-Day Decision Gate + At day 90, this venture is evaluated against these criteria: + - Minimum viable: [specific metric -- traffic, revenue, content volume] + - If minimum viable is NOT met: [specific action -- pause, pivot, or close] + - If minimum viable IS met: [specific next investment or scale decision] + + ## Final Recommendation + **FUND** / **PILOT FIRST** / **DO NOT FUND** + + Rationale: [2-3 sentences. Ground the recommendation in the data above. + If FUND: what makes the risk acceptable? + If PILOT FIRST: what small experiment resolves the uncertainty? + If DO NOT FUND: what fundamental barrier makes this a poor investment?] + + --- + + ## Sources + [All URLs cited above, numbered, with domain. + Format: N. [Title](URL) -- domain.com + Minimum 10 sources for a FUND recommendation.] + + - type: document + source_step: business_plan_body + dest_path: "deliverables/business-plans/business-plan-{task.id}.md" + commit_msg: "research: business plan task={task.id}" + + - type: close + rag_update: true + + - type: reply + target: channel + channel_name: "crimson_leaf:live-feed" + hint: | + Write a 3-4 line summary of the business plan just completed. + Include: opportunity name, final recommendation (FUND/PILOT FIRST/DO NOT FUND), + the projected 12-month revenue range (conservative to optimistic), + and one key risk or insight from the risk register. + Example: "Business plan complete: [name] -- FUND. 12-month projection $X-$Y. + Key risk: [risk] -- mitigated by [approach]." + +adjudication: + enabled: true + pass_threshold: 75 + deliverable_type: plan + criteria: + financial_realism: + weight: 40 + description: "Revenue projections state explicit assumptions -- no round numbers without basis. Break-even math is shown." + citation_density: + weight: 35 + description: "At least 10 real URLs cited inline. Every benchmark cites a source or is labeled as an estimate." + actionability: + weight: 25 + description: "90-day plan has specific tasks and decision thresholds, not vague goals. FUND/PILOT/PASS is grounded in data." diff --git a/templates/market_intelligence.yml b/templates/market_intelligence.yml index 4ac3c9c..a48fbb3 100644 --- a/templates/market_intelligence.yml +++ b/templates/market_intelligence.yml @@ -1,5 +1,5 @@ name: market_intelligence -description: "Orion researches a specific market opportunity with web search and produces a structured analysis document committed to Gitea. Spawned by opportunity_scan." +description: "Orion researches a specific market opportunity with 5-pass web research and produces a cited analysis document. Spawned by opportunity_scan." debug: true model: power system: agent_prompt @@ -22,17 +22,17 @@ steps: The market opportunity to research is described in the task message above. Formulate ONE precise web search query to find: - - Market size and growth rate for this niche - - Who the key players are and what they offer + - Total Addressable Market size and growth rate for this niche + - Industry reports, analyst data, credible statistics - Output ONLY the search query. Nothing else. + Output ONLY the search query. Nothing else. No preamble, no labels. - type: tool capability: Tool_WebSearcher input_from: search_query_1 mode: research - max_results: 6 - fetch_pages: 3 + max_results: 8 + fetch_pages: 4 optional: true output_key: market_data_1 @@ -40,10 +40,10 @@ steps: max_tokens: 200 output_key: search_query_2 hint: | - Your first search covered market size and key players. - Now formulate ONE focused query to find: - - Successful products in this niche, what they charge, who buys them - - Consumer pain points or unmet needs documented online + Pass 2 -- Competitive landscape. + Formulate ONE query to find the top 5 competitors in this niche: + - What they sell, what they charge, who buys from them + - Reviews, feature comparisons, pricing pages Output ONLY the search query. Nothing else. @@ -51,72 +51,171 @@ steps: capability: Tool_WebSearcher input_from: search_query_2 mode: research - max_results: 5 - fetch_pages: 2 + max_results: 8 + fetch_pages: 3 optional: true output_key: market_data_2 - type: think - max_tokens: 6000 + max_tokens: 200 + output_key: search_query_3 + hint: | + Pass 3 -- News and recent developments. + Formulate ONE query to surface what has happened in this niche + in the last 30-60 days. Look for: + - New products, launches, regulatory changes + - Funding events, acquisitions, market shifts + - "Why now" signals that make this a timely opportunity + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: search_query_3 + mode: news + max_results: 8 + optional: true + output_key: market_data_3 + + - type: think + max_tokens: 200 + output_key: search_query_4 + hint: | + Pass 4 -- Community voice and buyer pain points. + Formulate ONE query to find what real buyers or users say about + this niche on Reddit, Quora, or forums. Append "site:reddit.com OR site:quora.com" + to force community results. + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: search_query_4 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: market_data_4 + + - type: think + max_tokens: 200 + output_key: search_query_5 + hint: | + Pass 5 -- Revenue and monetization models. + Formulate ONE query to find how businesses in this niche make money: + - Subscription vs one-time vs affiliate vs advertising + - Actual revenue numbers if public, pricing benchmarks + - What the most profitable players do differently + + Output ONLY the search query. Nothing else. + + - type: tool + capability: Tool_WebSearcher + input_from: search_query_5 + mode: research + max_results: 6 + fetch_pages: 2 + optional: true + output_key: market_data_5 + + - type: think + max_tokens: 12000 output_key: market_analysis_body hint: | You are Orion, Director of Research at Crimson Leaf Holdings. - Produce a structured market analysis document for the opportunity in the task message. + Produce a research-grade market analysis document for the opportunity in the task message. - === PRIMARY RESEARCH (market size / players) === + === PASS 1: MARKET SIZE AND TAM === {market_data_1} - === END PRIMARY RESEARCH === + === END PASS 1 === - === SECONDARY RESEARCH (products / gaps / pricing) === + === PASS 2: COMPETITIVE LANDSCAPE AND PRICING === {market_data_2} - === END SECONDARY RESEARCH === + === END PASS 2 === - Cite every data point with [Title](URL) inline. - CITATION RULES: - - ONLY cite real URLs from the research data above (the [Title](URL) links in the research results). - - Do NOT invent URLs, placeholders, or use "project.rag" as a source. - - If a section has no real research data, write "No web data found. Structural analysis:" and continue. - - Every statistic must be followed by its source [Title](URL). + === PASS 3: NEWS AND RECENT DEVELOPMENTS === + {market_data_3} + === END PASS 3 === - Structure the document exactly as follows: + === PASS 4: COMMUNITY VOICE AND BUYER PAIN POINTS === + {market_data_4} + === END PASS 4 === + + === PASS 5: REVENUE AND MONETIZATION MODELS === + {market_data_5} + === END PASS 5 === + + CITATION RULES -- READ CAREFULLY: + - Every statistic must be followed immediately by [Title](URL) inline citation. + - Use ONLY real URLs found in the research passes above. + - Do NOT invent URLs, use placeholders, or cite "project.rag" as a source. + - If a section has no real web data, write: "(estimate - no web source found)" + and state the assumption explicitly. + - Target at least 8 real cited URLs across the full document. + + Structure the document EXACTLY as follows. Do not skip sections. # Market Analysis: [opportunity name from task message] --- ## 1. Market Overview - - Total Addressable Market: [size, growth rate, source] - - Key drivers: [2-3 forces growing this market] - - Key risks: [2-3 risks or headwinds] + - Total Addressable Market: [size, growth rate -- cite source] + - Key market drivers: [2-3 structural forces, each cited if possible] + - Key risks and headwinds: [2-3 specific risks] - ## 2. Target Audience - - Primary buyer: [specific demographic or job role] - - Buyer behavior: [how they discover and purchase in this niche] - - Pain points: [what they cannot get today] + ## 2. Recent Developments + [What happened in this niche in the last 30-60 days from Pass 3 news data. + If no news found, state "No significant developments found in the news pass." + Cite each item with [Title](URL).] - ## 3. Competitive Landscape - [Top 3-5 players: name | product | price | audience | strengths | weaknesses] - - White space: [what none of them do well that CL could own] + ## 3. Target Audience + - Primary buyer: [specific demographic, job role, or persona] + - Buyer behavior: [how they discover, evaluate, and purchase in this niche] + - Pain points: [what they cannot get today -- cite community sources from Pass 4] - ## 4. Crimson Leaf Positioning + ## 4. Competitive Landscape + Top competitors from Pass 2: + | Name | Product | Price | Audience | Strengths | Weaknesses | + |------|---------|-------|----------|-----------|------------| + [fill in table -- cite competitor URLs] + + White space: [what none of them do well that Crimson Leaf could own] + + ## 5. Monetization Models + [How does the money flow in this niche -- from Pass 5. + Subscription vs one-time vs affiliate vs advertising. + Include specific revenue figures or rates if found. Cite each.] + + ## 6. Crimson Leaf Positioning - Recommended entry point: [specific niche within the broader market] - - Differentiation: [how CL would be distinct from existing players] + - Differentiation: [how CL would be distinct, leveraging AI-speed and cost advantage] - Time-to-market: [realistic estimate given current CL capabilities] - ## 5. Revenue Potential - - Conservative (12-month): [number with assumptions] - - Base case (12-month): [number with assumptions] - - Optimistic (12-month): [number with assumptions] + ## 7. Revenue Projections (12 months) + - Conservative: [number with stated assumptions] + - Base case: [number with stated assumptions] + - Optimistic: [number with stated assumptions] + Note: If revenue data is weak, state assumptions explicitly rather than guessing. - ## 6. Recommendation - PROCEED / INVESTIGATE FURTHER / PASS + ## 8. Recommendation + **PROCEED** / **INVESTIGATE FURTHER** / **PASS** - Rationale: [2-3 sentences backing the recommendation with data from above] + Rationale: [2-3 sentences grounded in data from above. + If PROCEED: what makes this the right time? + If INVESTIGATE FURTHER: what specific question must be answered first? + If PASS: what structural barrier makes this unviable?] --- ## Sources - [List all cited URLs with titles] + [List every URL cited above, numbered, with title and domain. + Format: N. [Title](URL) -- domain.com] + + --- + + BUSINESS_PLAN_FLAG: [write "BUSINESS_PLAN_NEEDED: true" if recommendation is PROCEED, + otherwise write "BUSINESS_PLAN_NEEDED: false"] - type: document source_step: market_analysis_body @@ -126,26 +225,47 @@ steps: - type: close rag_update: true + - type: tool + action: enqueue_strategy + optional: true + condition: "BUSINESS_PLAN_NEEDED: true" + params: + company_slug: "crimson_leaf" + project_slug: "incubation" + task_type: "business_plan" + content: "Develop a full business plan for the PROCEED opportunity identified in market analysis {task.id}. The full market analysis is in the deliverables folder. Use it as your foundation -- do not repeat research already done." + agent_name: "orion" + + - type: tool + action: enqueue_strategy + optional: true + params: + company_slug: "crimson_leaf" + project_slug: "incubation" + task_type: "opportunity_scan" + content: "Follow-up scan triggered by completed market analysis {task.id}. Prioritize opportunities adjacent to the niche just researched." + agent_name: "orion" + - type: reply target: channel channel_name: "crimson_leaf:live-feed" hint: | Write a 2-3 line summary of the market analysis just completed. Include: topic researched, recommendation (PROCEED/INVESTIGATE FURTHER/PASS), - and one key data point with citation. - Example: "Market analysis complete: [topic] -- PROCEED. [Key finding with [Title](URL)]." + one key data point with citation, and whether a full business plan has been queued. + Example: "Market analysis complete: [topic] -- PROCEED. [Key finding [Title](URL)]. Business plan queued." adjudication: enabled: true pass_threshold: 70 deliverable_type: analysis criteria: - depth: - weight: 40 - description: "All 6 sections covered with specific data points -- no empty sections" - accuracy: + citation_density: weight: 35 - description: "Claims backed by research citations or clearly labeled as estimates -- no invented statistics" + description: "At least 8 real URLs cited inline -- no invented URLs or placeholders" + section_completeness: + weight: 35 + description: "All 8 sections present with substantive data -- no empty sections" actionability: - weight: 25 - description: "Recommendation (PROCEED/INVESTIGATE FURTHER/PASS) is specific and grounded in the analysis" + weight: 30 + description: "PROCEED/INVESTIGATE FURTHER/PASS recommendation is specific and grounded in cited data" diff --git a/templates/opportunity_scan.yml b/templates/opportunity_scan.yml index 681868e..894584e 100644 --- a/templates/opportunity_scan.yml +++ b/templates/opportunity_scan.yml @@ -1,5 +1,5 @@ name: opportunity_scan -description: "Crimson Leaf opportunity research -- Orion hunts trends and white-space every 6 hours." +description: "Crimson Leaf opportunity research -- Orion runs 4 targeted searches every hour: KDP charts, trending news, AI content market, and affiliate/review site opportunities." debug: true model: power system: agent_prompt @@ -11,98 +11,138 @@ sections: - agent - project - rag - - prior_results - message - instructions -builders: - prior_results: | - *** CURRENT PORTFOLIO *** - {steps[0].text} - - *** WEB RESEARCH RESULTS *** - {steps[2].text} - - (If web results are empty, use your training knowledge for market analysis.) - steps: - type: tool action: get_portfolio_report output_key: portfolio_data - - type: think - max_tokens: 2000 - hint: | - You are Orion, Director of Research at Crimson Leaf Holdings. - - Your job is to find the next big opportunity. Think about: - - What content formats are trending? (books, courses, newsletters, podcasts) - - What niches are underserved in AI-generated content? - - What seasonal trends are coming up? - - Formulate the single best search query to find market opportunities. - On the LAST LINE, write exactly: - SEARCH QUERY: [your query] - + # --- Search 1: Amazon KDP bestseller charts --- - type: tool capability: Tool_WebSearcher - input_from: last_text + input: "Amazon KDP Kindle bestseller list romance fiction top 100 this week new releases" + mode: research + max_results: 8 + fetch_pages: 3 optional: true + output_key: kdp_trends + + # --- Search 2: Trending news topics with content potential --- + - type: tool + capability: Tool_WebSearcher + input: "trending news topics viral stories this week content niche opportunities 2025" + mode: news + max_results: 8 + optional: true + output_key: news_trends + + # --- Search 3: AI content and publishing market trends --- + - type: tool + capability: Tool_WebSearcher + input: "AI content writing publishing trends 2025 self publishing income opportunities Substack newsletter" + mode: research + max_results: 8 + fetch_pages: 2 + optional: true + output_key: ai_content_trends + + # --- Search 4: Amazon affiliate and review site opportunities --- + - type: tool + capability: Tool_WebSearcher + input: "Amazon Associates affiliate review site high commission products niche 2025 best categories to review" + mode: research + max_results: 8 + fetch_pages: 3 + optional: true + output_key: affiliate_trends - type: think max_tokens: 6000 output_key: opportunity_analysis hint: | - You are Orion, the Visionary. You have the portfolio report and web research above. + You are Orion, Director of Research at Crimson Leaf Holdings. + You have four intelligence feeds below. Synthesize them into a ranked opportunity list. - Cross-reference what CL already produces with what the market wants. - Think like Elon Musk -- first principles, not incrementalism. + === PORTFOLIO (what CL already has) === + {portfolio_data} - Structure your analysis as: - 1. TOP 3 OPPORTUNITIES -- ranked by revenue potential vs. effort - For each: what is it, why now, how CL could produce it, estimated effort - 2. PORTFOLIO GAPS -- what are our subsidiaries NOT doing that they should be? - 3. ACTION ITEMS -- specific things David needs to do + === KDP BESTSELLER SIGNALS === + {kdp_trends} + (If empty: use your knowledge of current romance/fiction charts.) - Format action items as: "David, [verb]: [details]" - Be bold. Don't hedge. If an opportunity is real, say so with conviction. + === NEWS AND VIRAL TRENDS === + {news_trends} + (If empty: use your knowledge of current cultural moment.) + + === AI CONTENT AND PUBLISHING MARKET === + {ai_content_trends} + + === AMAZON AFFILIATE AND REVIEW SITE OPPORTUNITIES === + {affiliate_trends} + (Note: Amazon Associates pays 1-10% commission on referred purchases. A review + site publishing AI-written product reviews with affiliate links is a distinct + revenue stream from book sales -- lower effort, passive income.) + + Evaluate every signal against what CL can actually BUILD today: + - CL has AI writing workers (books, articles, reviews) + - CL has web search capability + - CL has Gitea for publishing content + - CL does NOT yet have a website, storefront, or affiliate account + + Structure your analysis: + + ### BOOK / CONTENT OPPORTUNITIES + Top 2 niches where CL should write books RIGHT NOW based on KDP charts or news trends. + For each: niche, why now (cite a signal), estimated monthly revenue if in top 100. + + ### AFFILIATE / REVIEW SITE OPPORTUNITY + Best product category for a CL review site (based on affiliate feed). + Consider: commission rate, search volume, competition, how AI-written reviews compete. + Be specific -- "kitchen gadgets under $30" beats "electronics". + Flag if this requires a new company (Crimson Leaf Reviews) or fits under an existing one. + + ### PORTFOLIO GAPS + What are our subsidiaries NOT doing that they should be doing this week? + + ### ACTION ITEMS + Format: "David, [verb]: [specific action]" + Include at least one action item about the affiliate opportunity if it looks viable. + + Be direct. Cite data from the feeds. Don't hedge. - type: think max_tokens: 600 output_key: opportunity_1_content hint: | - From the opportunity analysis above, extract the SINGLE TOP opportunity. + From the opportunity analysis above, extract the TOP BOOK/CONTENT opportunity. - Write a complete market_intelligence task brief for it. This brief will be - sent as the task message to Orion for deep research. + Write a market_intelligence task brief: + - What the opportunity is (1-2 sentences, cite the signal) + - Why now (timing -- what just happened in the market) + - Research questions: What does the KDP top 100 look like in this niche? + Who are the top 5 authors and what do they charge? What tropes are selling? + - CL entry angle: which agent writes this, which project slug to use - The brief must include: - - What the opportunity is (1-2 sentences) - - Why now (1 sentence on timing/signal) - - What market data to find (2-3 specific questions to answer) - - What CL's potential entry angle might be (1 sentence) - - If the analysis found NO compelling new opportunity (all were already covered - in the portfolio or too vague), output exactly: none - - Output ONLY the brief text or the word none. No preamble. + If no compelling book opportunity was found, output exactly: none + Output ONLY the brief or the word none. No preamble. - type: think max_tokens: 600 output_key: opportunity_2_content hint: | - From the opportunity analysis above, extract the SECOND-BEST opportunity - (not the same as opportunity 1). + From the opportunity analysis above, extract the AFFILIATE/REVIEW SITE opportunity. - Write a complete market_intelligence task brief for it using the same format: - - What the opportunity is - - Why now - - What market data to find - - CL's potential entry angle + Write an affiliate_intelligence task brief: + - What the product category is and why it was selected + - Amazon Associates commission rate for this category if found in the research + - Key questions to answer: Who are the top review sites in this niche? + What content gaps exist? How many articles needed for $500/month in commissions? + - What Crimson Leaf would need to build: domain, content volume, affiliate account - If there is no distinct second opportunity worth researching, output exactly: none - - Output ONLY the brief text or the word none. No preamble. + If no compelling affiliate opportunity was found, output exactly: none + Output ONLY the brief or the word none. No preamble. - type: tool action: enqueue_strategy @@ -119,34 +159,35 @@ steps: params: company_slug: "crimson_leaf" project_slug: "incubation" - task_type: "market_intelligence" + task_type: "affiliate_intelligence" content: "{opportunity_2_content}" - type: think max_tokens: 100 output_key: dispatch_summary hint: | - Count how many research tasks were commissioned this cycle. Look at opportunity_1_content and opportunity_2_content above. - If both are not "none": output "Commissioned 2 research tasks." - If only opportunity_1_content is not "none": output "Commissioned 1 research task." - If both are "none": output "No research tasks commissioned -- no new opportunities found." + If both are not "none": output "Commissioned 2 research tasks (market_intelligence + affiliate_intelligence)." + If only opportunity_1_content is not "none": output "Commissioned 1 research task (book opportunity -- market_intelligence)." + If only opportunity_2_content is not "none": output "Commissioned 1 research task (affiliate opportunity -- affiliate_intelligence)." + If both are "none": output "No research tasks commissioned -- no new signals found." Output ONLY that sentence. - type: reply target: channel channel_name: "crimson_leaf:live-feed" hint: | - Write a 3-5 line opportunity scan summary for Orion's report. + Write a 4-6 line opportunity scan summary. Start with: "Opportunity scan complete. {dispatch_summary}" - Then summarize: - - The top opportunity found (or "no new opportunities" if none) - - One key market signal with a citation if available + Then cover: + - Top book/content signal (niche + why now, cite source if available) + - Top affiliate/review site signal (category + commission rate if found) - One portfolio gap identified + - One action item for David - Be concise and action-oriented. + Be specific. Cite real data from the research. No filler. adjudication: enabled: false diff --git a/templates/pulse_check.yml b/templates/pulse_check.yml index 317a63a..55e1d2e 100644 --- a/templates/pulse_check.yml +++ b/templates/pulse_check.yml @@ -65,7 +65,9 @@ steps: hint: | Write a 2-line status update for Discord. Line 1: "Pulse [HEALTHY/ATTENTION/IDLE] -- projects reviewed" - Line 2: one sentence noting anything flagged (failures, idle projects, workers down) or "All systems nominal." + Line 2: one sentence noting anything flagged (failures, idle projects, workers down, + cost anomalies -- flag if total_spend_usd_today > $1.00 or any company + cost_per_task_usd_30d > $5.00) or "All systems nominal." Use data from the portfolio report. 2 lines max. No extra commentary. - type: reply diff --git a/templates/strategic_review.yml b/templates/strategic_review.yml index 29825f9..da07aae 100644 --- a/templates/strategic_review.yml +++ b/templates/strategic_review.yml @@ -76,7 +76,9 @@ steps: - If you are Edgar (The Operator): Focus on execution health and metrics only. Are companies executing? Are we moving toward the mission success criteria? - Report: CLP chapter velocity, CLM campaign status, adjudication pass rates. + Report: CLP chapter velocity, CLM campaign status, adjudication pass rates, + and total_spend_usd_30d from the portfolio report. If cost is up vs last cycle, + flag it. If cost_per_polished_chapter_usd_30d > $0.50, flag it. DO NOT discuss what company to propose -- proposals are handled by a separate gate after this debate. The mission charter already defines what gaps exist and what companies are authorized. Your job here is execution reporting. @@ -105,9 +107,11 @@ steps: Structure: 1. EXECUTIVE SUMMARY -- one paragraph on portfolio health and mission progress - 2. MISSION GAP AUDIT -- assess each gap from the mission charter. Is Gap 1 (CLO) - still the top priority? Has anything changed that elevates a different gap? - 3. CRITICAL ISSUES -- things needing immediate attention + 2. MISSION GAP AUDIT -- assess each gap from the mission charter. Is Gap 3 (KDP + pipeline) now the top priority since Gap 1 (CLO) is resolved? Has revenue + arrived yet? Cite total_royalty_usd_30d from the portfolio report if available. + 3. CRITICAL ISSUES -- things needing immediate attention (include cost overruns: + if total_spend_usd_30d > $20.00 or cost_per_polished_chapter > $0.50, flag it) 4. OPPORTUNITIES -- ranked by priority 5. ACTION ITEMS -- specific tasks for David: "David, [verb]: [details]"