spec

Lantronix UAS Research Workbench Phase 1

2026-04-22

Lantronix UAS Research Workbench Phase 1

Build phase 1 as an internal research workbench inside OpenClaw and Mission Control. The core decision is to optimize for three consulting outputs, competitive landscape, competitive positioning with technical differentiation, and target customer segments, using a structured US-first dataset built around companies, products, customer segments, and evidence. Do not build a standalone app yet.

1. Problem Statement and Goal

Pete needs a reliable way to produce Lantronix deliverables without missing important competitors, adjacent players, products, or customer opportunities. The current UAS work has real value, but it started from an arbitrary first source set and was shaped partly by a broader product idea in UAS Strategic Intelligence Platform. That creates two problems:

The goal of phase 1 is simple: create an internal Lantronix research workbench that lets Pete produce credible, repeatable strategy outputs for Lantronix leadership, especially the SVP Engineering, VP Product, and EVP Product & Strategy, while keeping the system narrow enough to improve fast.

2. Success Metric

Phase 1 is successful when all of the following are true:

3. Current State

What exists today:

What is missing:

Current task fit, blunt version:

TaskCurrent fitRecommendation
JAM-15right parent, wrong level of abstractionkeep and repoint to this phase-1 plan
JAM-26directionally useful but too XPO-centrickeep, rewrite around taxonomy + recast + targeted gap fill
JAM-24useful only if model quality blocks progressdefer or archive for now
JAM-27out of phase for US-first workarchive for now
JAM-29premature until workflow is provenarchive for now

4. Platform Capabilities

OpenClaw and the current Mission Control setup already support most of what phase 1 needs:

What OpenClaw does not require for phase 1:

That is why OpenClaw and Mission Control are the right substrate for phase 1. The work is still analyst-driven, source-heavy, and evolving.

5. Community Patterns

The proven pattern for early market-intelligence systems is internal-first: structured records, analyst judgment, human-in-the-loop updates, and templated outputs. Teams often start with flexible systems like Airtable, Notion databases, or internal research stacks before they earn the cost and rigidity of a dedicated product. Closed research products like AlphaSense are useful benchmarks for output quality, but they are the wrong build target for phase 1 because the real need here is custom industry structure and consulting leverage, not a general-purpose market-intel product.

Within the OpenClaw world, the closest proven pattern is internal skill-driven workflow plus artifact delivery, not a standalone app. ClawHub and the current workspace setup both point toward narrow internal workflows first.

6. Options

OptionDescriptionComplexityToken CostReliabilityMaintenance burdenVerdict
A. Standalone app nowBuild a dedicated Lantronix/UAS app with its own data model, UI, and workflowHighHighMediumHighToo early
B. Internal research workbench in OpenClaw/Mission ControlUse local data files, artifacts, light internal views, and targeted workflow automationMediumMediumHighMediumBest phase-1 path
C. Spreadsheet-only workflowKeep this as Sheets plus notes with light manual synthesisLowLowMediumMedium-high human overheadFine as fallback, weaker for scale and traceability

Option A: Standalone app now

Pros:

Cons:

Option B: Internal research workbench in OpenClaw/Mission Control

Pros:

Cons:

Option C: Spreadsheet-only workflow

Pros:

Cons:

7. Recommendation

Use Option B: an internal research workbench inside OpenClaw and Mission Control.

Key decisions:

This is the right level of ambition. It gets Pete a useful system fast and leaves room to widen later if the Lantronix engagement proves it deserves more.

8. Security Considerations

9. Implementation Scope

What David should build after approval:

Data model

Create or migrate to a phase-1 data structure with these first-class record types:

Minimum fields

Company

Product

Customer Segment

Evidence / Source

Workflow scope

Task restructuring proposal

After Pete reviews this spec:

10. Validation Criteria

The build is good enough for phase 1 when these tests pass:

  1. Recast test: the existing 125-company set can be reclassified into the new taxonomy without major field ambiguity on the important categories.
  2. Gap test: the system can show which competitor classes, product classes, and customer segment classes are undercovered.
  3. Filter test: Pete can filter and sort the system by core fields needed for the first three report types, using the simplified relevance flag for default reporting.
  4. Traceability test: each important positioning claim in a draft report points to at least one evidence item.
  5. Report test: Pete can generate first-draft versions of the three target report types from the system.
  6. Kickoff-prep test: the system helps Pete ask better Lantronix questions before it tries to answer everything.

11. Category

Skill

This is a multi-step internal workflow for building, maintaining, and using a structured consulting research system.

12. Context Loading

When this workflow is active, load the following context:

Always load for Lantronix UAS strategy work

Load for current dataset and taxonomy work

Load when historical framing is useful

Trigger conditions

13. Guardrails

14. Handoff

Deliver this plan to Pete as a Mission Control artifact using the Tailscale artifact URL:

http://vinnys-mac-mini.tail31784c.ts.net:3100/artifacts/spec_lantronix_uas_research_workbench_phase1

After Pete reviews the artifact:

  1. update JAM-15 to link to this artifact as the active spec
  2. move JAM-15 to needs-review
  3. keep assignee on Pete
  4. after approval, rewrite or archive the stale UAS backlog tasks before any build work starts