Opportunity Backlog
> Purpose: This backlog tracks ideas by _strategic_ leverage — long-term authority, evergreen value, unique positioning. It complements the Product Backlog which tracks product/engineering opportunities. Both are maintained independently.
>
> Ranking heuristic: Impact × Originality × Compounding potential / Effort
---
Intelligence Update — 2026-07-13 (Trigger: Commenting UX wishlist update + competitive research completion)
Signal: What I want for the commenting experience to be was updated by Horacio's Gardener. The document now includes 5 specific UX requests and a complete body of cross-referenced competitive research (Slite, Coda, Linear, Google Docs inline comment patterns).
Key observations:
External research corpus is complete. The related links section now catalogs inline comment UX from 4 major platforms + Seed's own pain point doc. This is a synthesized research corpus — not raw notes. An editorial synthesis is low-effort and high-leverage.
Request #2 (block-level comments as first-class blocks) is a genuinely novel positioning. No platform treats individual comments as block-addressable, range-linkable primitives. This is the Seed differentiator for collaborative documents. It connects directly to Signature Idea #3 (documents as first-class citizens).
Requests #3–4 (mobile comments) are product-level, not editorial. The mobile experience gaps (janky drawer, laggy editor) are engineering problems. But documenting the mobile-first approach to hypermedia commenting is editorial leverage — especially for mobile web developers evaluating Seed.
Request #5 (no redirect) is explicitly linked to Remodeling the web. This creates a thematic bridge between the commenting UX and the broader web remodeling thesis.
Implication for rankings: Adding 3 new commenting UX opportunities (#6, #8, #10). Existing #1 (Manifesto) and #2 (Vibe Code Bridge) remain unchanged — they address a 10x larger market. The commenting opportunities are Seed-specific brand-building, lower-effort, and slot into Q4 2026.
---
Current Rank (2026-07-13)
| Rank | Opportunity | Impact | Effort | Confidence | Compounding? | Why now |
|------|------------|--------|--------|-----------|-------------|---------|
| 1 | **State Machine Manifesto for LLM Agents** | Very High | Medium | 9/10 | Yes — positions all other content | Cluster fully connected; cross-links complete; no canonical argument exists |
| 2 | **"From Vibe Code to Deterministic Agents"** | High | Low | 9/10 | Yes — captures mainstream search traffic | "Vibe code like a PRO" connects trend to thesis |
| 3 | **"What Goes Wrong Without State Machines"** | High | Low | 8/10 | Yes — strengthens positive argument | Unwritten counterpoint; makes the manifesto land harder |
| 4 | **Linear→Seed Agent Pipeline** | High | Medium | 8/10 | Yes — proves the pipeline as product demo | Community MCP guide exists; cross-links drive traffic to architecture docs |
| 5 | **Agent Determinism Workshop/Talk** | High | Medium | 8/10 | Yes — distribution channel for OSS | Conference calls opening for 2026 |
| 6 | **NEW: "What Block-Level Commenting Unlocks"** | Medium-High | Low | 8/10 | Yes — positions Seed's unique differentiator | Competitive research complete; no one has written this |
| 7 | **Cross-Account Knowledge Graph Curation Playbook** | Medium | Low | 7/10 | Yes — enables automated curation at scale | Gardener demonstrated cross-link automation; product-funnel pattern emerged |
| 8 | **NEW: "What We Learned from Slite, Coda, Linear, and Google Docs Comments"** | Medium | Low | 7/10 | Moderate — authoritative reference | Research complete; synthesis requires only writing |
| 9 | **Seed Agent SDK Documentation Site** | High | Medium | 7/10 | Yes — official docs compound long-term | Cross-links funnel users from practical content to architecture docs |
| 10 | **NEW: "Commenting That Works on Mobile First"** | Medium | Low | 6/10 | Moderate — mobile dev audience | Specific pain points articulated; fix-and-tell content pattern |---
1. State Machine Manifesto for LLM Agents
_Unchanged from previous version._ See full analysis on Content Roadmap.
---
2. "From Vibe Code to Deterministic Agents" — The Framing Bridge
_Unchanged from previous version._
---
3. "What Goes Wrong Without State Machines" — Negative-Space Analysis
_Unchanged from previous version._
---
4. Linear→Seed Agent Pipeline — Reference Implementation
_Unchanged from previous version._
---
5. Agent Determinism Workshop/Talk
_Unchanged from previous version._
---
6. [NEW] "What Block-Level Commenting Unlocks" — Seed's Commenting Differentiator
Thesis: Seed Hypermedia treats comments as block-addressable, linkable, range-selectable primitives — the same way it treats document blocks. This is a genuinely novel approach that no existing platform (Slite, Coda, Notion, Google Docs, Linear) implements. It turns discussions into structured, linkable knowledge — not ephemeral threads.
Evidence: What I want for the commenting experience to be request #2. Competitive research corpus (Slite, Coda, Linear, Google Docs) from the same doc shows no platform does this. Discussions Panel Is Confusing.
Impact: Medium-High — positions Seed as the only platform with block-addressable comments. Narrows comparison gap vs. Notion/Coda.
Effort: Low — research complete. 1,000–1,500 word comparative analysis.
Confidence: 8/10 — defensible; competitive gap is real and documented.
Concrete next action: Write the piece — start with a comparison table of how each platform handles comment addressing, then explain Seed's block-level approach.
---
7. Cross-Account Knowledge Graph Curation Playbook
Observation: The Gardener agent's automated cross-linking of How to Background Agents across 3 accounts proves this works. The Product Backlog cross-link creates a product-funnel pattern.
Format: Internal playbook (1,000 words). Could evolve into a Seed Hypermedia case study.
Concrete next action: Analyze the pattern: trigger → context read → cross-link insert → product-funnel connect.
---
8. [NEW] "What We Learned from Studying Slite, Coda, Linear, and Google Docs Comments" — Comparative UX Analysis
Thesis: A synthesis of the 4-platform competitive research corpus. Each platform solves inline commenting differently. The analysis extracts what works, what doesn't, and what's missing — positioning Seed's approach as the next evolution.
Evidence: Complete corpus in What I want for the commenting experience related links: Slite, Coda, Linear, Google Docs.
Impact: Medium — authoritative reference content. Seeds future comparison pieces and positions Seed as thoughtful on collaboration UX.
Effort: Low — research done. 1,200–1,800 words + simple diagrams.
Confidence: 7/10 — useful but not a differentiator piece; Slite/Coda audience may not overlap with Seed's.
Concrete next action: Extract patterns (margin annotations, inline highlights, threaded replies) and evaluate each platform. Identify gaps that Seed fills.
---
9. Seed Agent SDK Documentation Site
_Unchanged from previous version._ See Content Roadmap for details.
---
Format: Overview, Quickstart, Concepts, State machine model, API reference.
Dependencies: Agent OSS packages (Opportunities 19–22) should be stable first.
Concrete next action: Map existing content to a docs outline. Identify gaps. Priority: Quickstart guide first.
10. [NEW] "Commenting That Works on Mobile First"
Thesis: The mobile commenting experience (#3, #4 in the wishlist) is explicitly broken in Seed today. Fix-and-tell is a proven content pattern. Mobile web developers evaluating Seed are a growing audience.
Evidence: What I want for the commenting experience requests #3 (janky drawer) and #4 (laggy editor on mobile).
Impact: Medium — mobile web developers evaluating Seed. Case study format.
Effort: Low — but depends on mobile fixes shipping first. Content can follow product.
Confidence: 6/10 — product dependency. Premature to write before fixes ship.
Concrete next action: Monitor Product Backlog for mobile commenting fix completion. Write as soon as the before/after story is available.
Do you like what you are reading? Subscribe to receive updates.
Unsubscribe anytime