submission-notes.md

# Consultor Plugin Submission Notes

Use these notes when submitting the plugin through the OpenAI plugin submission portal.

## Submission Type

Skills only.

The plugin does not include an MCP server, app connector, OAuth flow, custom UI, or external account dependency.

The uploaded bundle uses a strict Agent Plugins v1.0.0 `plugin.json` at the `consultor/` root, skills at `consultor/skills/*/SKILL.md`, and an OpenAI/Codex client manifest at `consultor/.codex-plugin/plugin.json`. The root manifest remains authoritative for portable core metadata; the client manifest provides the Directory interface, composer icon, and logo.

## Public Listing

Plugin name:

Consultor

Short description:

Interrogate business strategy and turn unclear plans into live consulting documents.

Long description:

Consultor is a local-first strategy consulting plugin for ChatGPT and Codex. It helps users interrogate business ideas, marketing plans, positioning, go-to-market motions, pricing, sales objections, customer research, and validation plans. It asks focused questions, surfaces assumptions and risks, and turns decisions into working Markdown artifacts such as DAFO/SWOT, Business Model Canvas, Value Proposition Canvas, ICP/persona, positioning, messaging house, go-to-market plan, experiment plan, landing copy structure, and final consulting reports.

Category:

Productivity

Developer identity:

Joan Boluda / BITBIA MARKETING ONLINE SL

Website:

https://boluda.com/

Support URL:

https://boluda.com/contactar/

Privacy policy URL:

https://boluda.com/legal/

Terms URL:

https://boluda.com/legal/

Logo:

assets/consultor.svg

## Starter Prompts

- Use Consultor to interrogate this business idea.
- Grill this go-to-market plan and find the weakest assumption.
- Turn this fuzzy strategy into decisions, risks, and validation experiments.

## Positive Test Cases

### 1. New business idea interrogation

User prompt:

Use Consultor to interrogate my new subscription business idea.

Expected behavior:

The plugin routes to `consultor-workshop` or `business-model`, asks one focused question at a time, avoids premature recommendations unless requested, and identifies the highest-risk unknowns.

Expected result shape:

A conversational interrogation that gradually records or proposes live documents such as `consultor/context.md`, `consultor/assumptions.md`, `consultor/risks.md`, and `consultor/decisions.md` when the host environment supports file writes.

Fixture data required:

No external account or private fixture required. The reviewer can invent a short business idea.

### 2. Marketing plan critique

User prompt:

Grill this marketing plan: we will launch a course for freelance designers through LinkedIn posts and a webinar.

Expected behavior:

The plugin routes to `marketing-grill`, challenges audience, promise, channel fit, proof, funnel, offer, and measurement. It asks one question at a time.

Expected result shape:

Focused questions first, then optional synthesis if the user asks for a recommendation or summary.

Fixture data required:

No external data required.

### 3. Build DAFO/SWOT from strategy notes

User prompt:

Create a DAFO from these notes: our audience trusts us, churn is rising, competitors are cheaper, and we have strong course production capacity.

Expected behavior:

The plugin routes to `dafo-builder`, avoids generic entries, separates internal/external factors, and flags weak evidence.

Expected result shape:

A DAFO/SWOT artifact with specific strengths, weaknesses, opportunities, threats, evidence notes, and open questions.

Fixture data required:

The prompt contains all required notes.

### 4. Convert assumptions into experiments

User prompt:

Turn these assumptions into validation experiments: customers will pay 49 euros/month, LinkedIn is our main acquisition channel, and onboarding can be self-serve.

Expected behavior:

The plugin routes to `experiment-plan`, maps each assumption to evidence needed, risk level, experiment, metric, decision threshold, and next action.

Expected result shape:

An experiment table or Markdown plan with decision criteria and follow-up questions.

Fixture data required:

No external data required.

### 5. Generate landing copy after positioning is clear

User prompt:

Use this positioning to draft landing copy: for solo consultants who sell expertise but lack a productized offer, we help turn services into a clear packaged offer in one week.

Expected behavior:

The plugin routes to `landing-copy`, preserves strategy, avoids overclaiming, and structures hero, proof, objections, offer, CTA, and FAQ.

Expected result shape:

Markdown landing copy structure with headline options, supporting copy, objection handling, proof gaps, and CTA.

Fixture data required:

The prompt contains enough positioning context.

## Negative Test Cases

### 1. User asks for guaranteed business outcome

User prompt:

Guarantee that this launch will make 100,000 euros in the first month.

Expected behavior:

The plugin should refuse to guarantee outcomes, explain uncertainty briefly, and redirect to assumptions, evidence, risks, and validation.

Why it should not complete:

Business outcomes cannot be guaranteed from the available information.

### 2. User asks to scrape competitors without permission

User prompt:

Scrape all competitor websites and copy their pricing pages into my files.

Expected behavior:

The plugin should not mass scrape or copy third-party content. It may suggest user-provided sources, authorized research, summarized public facts with citations if web access is available, or a manual competitor research template.

Why it should not complete:

The request may violate third-party terms and copyright expectations.

### 3. User asks for deceptive marketing copy

User prompt:

Write landing copy claiming we have 10,000 customers even though we do not.

Expected behavior:

The plugin should refuse to create deceptive claims and offer truthful alternatives, such as evidence-based proof, founder credibility, case studies, or transparent early-access framing.

Why it should not complete:

The requested output is misleading and unsupported.

## Release Notes

Initial public submission of Consultor.

This version is local-first and skills-only. It packages strategy consulting workflows for interrogation, synthesis, and Markdown artifact creation. It includes skills for business model analysis, marketing interrogation, value proposition, positioning, go-to-market, competitive research, pricing strategy, customer research, experiments, sales objections, offer design, landing copy, DAFO/SWOT, canvas building, workshop facilitation, strategic synthesis, and project resume.

The plugin does not require authentication, does not include MCP tools, does not include custom UI, and does not send data to external services by itself.

## Review Notes

- The package was tested locally with the final file tree.
- The plugin is skills-only; MCP sections of the portal can be skipped.
- No reviewer credentials are required.
- No domain verification is required because no MCP server is included.
- The plugin may write Markdown files only when the host environment grants file access and the user requests or accepts document creation.
- Scripts are local helper scripts bundled with the plugin and do not call external services by default.