Positioning Framework Selection — gblock-party
Skill: /positioning (Mode A — Framework Selection)
Product: gblock-party — always-on agent workstation
Date: 2026-06-01
Mode: Market Positioning (pre-product, no customer feedback)
Select positioning frameworks to execute. Each framework produces an intermediate research artifact. After all run, /positioning --synthesize combines them into research/positioning.md.
Table of Contents
Mode Detection
Detected: Market-positioning mode.
Evidence for market mode:
- Missing
research/customer-feedback.md— no post-launch customer evidence exists - No production codebase with live users detected — gblock-party is pre-product
- User invocation did not contain "product", "post-launch", or "obviously-awesome"
This means we use market-positioning frameworks that build hypothesized positioning from ICP research, competitive analysis, and journey mapping — not customer-grounded frameworks like Obviously Awesome.
Available Evidence Summary
research/icp.md Ready
Primary ICP: Power Users / Solo Founders running 2-10+ AI coding agents on remote infrastructure. 4 ICPs total with cross-ICP prioritization. Pain map, trigger events, value drivers, market sizing ($12.8B TAM, ~$2-8M SOM), and discovery/evaluation behavior documented.
Key positioning inputs: Value language ("close the laptop, pick up where I left off", "see all agents in one view", "approve from my phone"), deal-breakers (vendor lock-in, price >$30/mo), the buyer=user=admin dynamic.
research/competitive-analysis.md Ready
27 competitors across 7 categories. No product combines BYO-client + managed infrastructure + multi-agent + web/mobile + zero-inbound security. Front-end UI layer is commoditizing (5+ OSS tools). Durable value is the managed infrastructure layer.
Key positioning inputs: Category claimed: "always-on agent workstation" (resolved via search term research), infrastructure thesis (UI is a client for the infra API), competitive differentiation axis (persistent managed infra vs. desktop apps that die when laptop closes).
research/journey-map.md Ready
Full lifecycle mapped. Two-track model (BYO-client + Agent Board). Aha moment: "started 5 agents, closed laptop, checked from phone — 3 PRs ready." 5 critical moments identified. Journey gaps: approval UX on mobile, session recap, review bottleneck tooling, churn defense.
Key positioning inputs: Aha moment definition, critical moment sequence (setup → persistence proof → phone proof → scale proof → value justification), two-track convergence on managed infrastructure.
research/customer-feedback.md Not Available
No customer feedback exists. This is expected for a pre-product project. Product-mode frameworks (Obviously Awesome) are disabled.
research/monetization.md Optional — Not Yet Run
No monetization research exists. Pricing signals available from ICP ($9-29/mo sweet spot) and competitive analysis (credit-based pricing hostile, tmux is free). Not required for positioning but would strengthen value perception claims.
Framework Selection
Based on market-positioning mode and the available evidence, these frameworks are recommended. Check the ones you want to execute:
Execution Plan
After approval, selected frameworks will be written to tasks/todo.md as sequential steps. You run /exec to execute each one. Each framework reads the same evidence base and produces an intermediate artifact under research/.
Proposed tasks/todo.md content (based on default selection):
## Positioning Framework Execution - [ ] Run `/positioning/frameworks/jtbd-positioning` — JTBD analysis → research/positioning-jtbd.md - [ ] Run `/positioning/frameworks/strategic-canvas` — Blue Ocean gap analysis → research/positioning-strategic-canvas.md - [ ] Run `/positioning/frameworks/moore-positioning` — Positioning hypothesis → research/positioning-moore.md - [ ] Synthesize: `/positioning --synthesize` — Combine framework outputs into research/positioning.md
If you also select Category Design, this step will be inserted before synthesis:
- [ ] Run `/positioning/frameworks/category-design` — Category creation analysis → research/positioning-category-design.md
Execution flow: Each /exec invocation picks the next unchecked item, enters plan mode for approval, then executes. Frameworks are independent and can be reordered, but the synthesis step must be last.
Evidence Matrix
| Positioning Input | Source | Evidence | Inference | Confidence | Feeds Framework |
|---|---|---|---|---|---|
| Primary job: "Close laptop, pick up from phone" | ICP, Journey Map | #1 stated desire across VPS+agent articles; aha moment in journey map | Session persistence + mobile access is the core value proposition, not the UI | High | JTBD, Moore |
| No product combines 5 key capabilities | Competitive Analysis | 27-competitor feature matrix; no overlap on BYO-client + managed infra + multi-agent + mobile + zero-inbound | Clear category whitespace for a combined solution | High | Canvas, Moore, Category |
| Front-end UI is commoditizing | Competitive Analysis | 5+ OSS agent dashboards (Multica, T3, amux, Claude Squad, Emdash) all free | Durable value is managed infrastructure, not the dashboard UI | High | Canvas, JTBD |
| Buyer = user = admin (solo decision) | ICP | Solo devs, budget authority is self, decision in minutes not weeks | Positioning must speak directly to the practitioner, not a buying committee | High | Moore, JTBD |
| "Not competing on terminal-in-browser" | ICP, Competitive Analysis | Anti-positioning signal from research; code-server and CDEs already cover browser terminals | Must position away from CDE/IDE category to avoid commoditized comparison | High | Canvas, Category |
| Price sensitivity: must be under $30/mo | ICP | DIY (tmux) is free; T3 Code, Emdash free; ICP sweet spot $9-29/mo | Value must be visceral — "I literally could not do this before" — to justify any cost over free | High | Moore, JTBD |
| First-party tools are single-agent only | Competitive Analysis | Claude Channels, Copilot Remote, Codex Mobile all single-provider | Cross-provider orchestration is the persistent gap; risk if vendors go multi-agent | Medium | Canvas, Moore |
| "Always-on agent workstation" as category name | Competitive Analysis, search term research | Resolved: "always-on" maps to user language, "workstation" unclaimed, "agent" avoids GPU infra confusion | Research-backed; pending direct user language testing | Medium-High | Category, Canvas, Moore |
Confidence & Assumption Register
| Assumption | Status | What Would Change It |
|---|---|---|
| Market mode is correct (no real customers yet) | Validated | Customer feedback or live user data emerging |
| JTBD + Canvas + Moore is sufficient for initial positioning | Provisional | Canvas reveals no existing category fits → add Category Design |
| Category Design not needed yet | Provisional | Strategic Canvas showing the product doesn't fit "developer tools" or "CDE" categories |
| Obviously Awesome is premature | Validated | Customer feedback with real post-launch evidence |
| Positioning evidence is sufficient to begin | Validated | ICP + competitive + journey all exist with high-confidence claims |
Approval Gates
Gate 1: Framework Selection Confirmation
Confirm which frameworks to execute. The checkboxes above control the selection — this gate confirms your intent.
Gate 2: Is the evidence base sufficient to begin positioning work?
Three research artifacts are available (ICP, competitive analysis, journey map). Monetization research is not yet done. Customer feedback does not exist (pre-product).
Gate 3: Proposed File Changes
After approval, the following file will be updated:
- Update:
tasks/todo.md— add positioning framework execution steps
No research files will be created until each framework runs via /exec.
Compile & Submit
Feedback (optional — send before answering all gates)
If any section needs revision, set the feedback controls above and compile feedback here.