Manage Client Discovery

What the company appears to be: Manage Client Discovery is a self-described Codex plugin that helps teams structure client interviews by turning scattered project context into evidence-grounded interview plans. It uses GPT-5.6 for interpretation and question prioritization, while maintaining explicit user control over source selection, data handling, and confirmation of new knowledge.

What changed: The author describes an evolution from traditional client discovery methods — which they claim are inefficient due to repeated questions or premature requirements gathering — toward a more structured, state-driven workflow that separates asynchronous document review from live conversation. This approach emphasizes preserving project evidence references and using human-in-the-loop systems to avoid AI-generated assumptions becoming project truth.

Single most important open question: Is there any evidence of real-world usage or adoption beyond the synthetic demo loop described? The description states no revenue, customers, or traction data are available; all claims are self-reported and unverified.

The description states that Manage Client Discovery is a Codex plugin. It contains one state-driven Skill, focused workflow references, and Python standard-library scripts for deterministic state changes. It uses GPT-5.6 to interpret evidence, classify uncertainty, discover gaps, and prioritize questions.

It produces two deliverables before a meeting:

1. A lightweight request for existing artifacts that can be reviewed asynchronously.
2. A timed live-interview plan focused on experience, judgment, tensions, exceptions, and tradeoffs.

After the interview, it extracts proposed facts, decisions, unresolved questions, and tasks from notes but does not silently rewrite project knowledge. The user reviews proposals first, and only confirmed learning becomes part of the saved project state.

The plugin also supports resumable workflows that inspect saved state, explain next steps, and ask for user decisions before continuing.

Evidence: Self-reported by author; no independent verification or demonstration of actual product functionality beyond a demo loop.

The description states that Manage Client Discovery was inspired by the need to improve client discovery processes. It positions itself as a safer alternative to traditional methods, which it claims often result in:

• Repeating already-answered questions
• Moving into requirements before understanding key unknowns

It aims to turn "scattered context" into "evidence-grounded interview plans", emphasizing:

• Preservation of evidence references and provenance
• Separation between pre-meeting document review and live conversation
• Human-in-the-loop confirmation gates to prevent AI-generated assumptions from becoming project truth

The author also notes that good discovery support is not mainly about generating more questions, but rather about deciding what should be learned asynchronously, what deserves live conversation, and how to treat assumptions as confirmed knowledge.

Inference: The evolution reflects a shift from generic AI-assisted tools toward structured workflows with explicit boundaries and human confirmation. However, this claim lacks evidence of real-world application or user feedback.

The description does not clearly define the target customer or ideal customer profile (ICP). It implies that the tool is intended for teams working on projects where client discovery is a challenge — particularly those dealing with distributed context across documents, trackers, and notes.

It suggests use cases involving:

• Project teams needing structured interviews
• Organizations looking to avoid redundant or premature requirements gathering
• Teams wanting to preserve project knowledge without AI assumptions becoming truth

There is no mention of specific industries, roles (e.g., product managers, designers), or team sizes beyond the single-member team mentioned in the project details.

Evidence: Not evidenced. No explicit customer segment or persona defined.

The description does not contain any information about pricing, monetization strategy, or business model. It only describes a plugin built for Codex and mentions that it was submitted to an OpenAI hackathon.

There is no indication of whether the tool will be sold, licensed, offered as part of a SaaS platform, or used internally by the author’s team.

Evidence: Not evidenced.

The project is built as an installable Codex plugin, using:

• GPT-5.6 for interpretation and question prioritization
• Human-in-the-loop systems to maintain control over data handling
• State-machine architecture with deterministic state changes
• Python standard-library scripts for workflow automation
• JSON and Markdown formats for artifact generation

It includes:

• One state-driven Skill
• Workflow references
• Regression and end-to-end tests (11 automated tests)
• Plugin validation, Skill validation, and fresh-session demo validation

The author notes that the plugin was designed to be resumable and no-argument, meaning it can start at any stage and adapt based on saved workflow state.

Evidence: Self-reported; no independent verification of technical implementation or delivery quality.

There is no evidence of traction, revenue, customers, or adoption beyond the synthetic demo loop described. The project was submitted to a hackathon, and the author states that post-submission work will refine conversational boundaries and explore future integrations.

The only maturity signal mentioned is:

• A complete synthetic demo loop
• Deterministic automated tests (11)
• Fresh-session validation using a local brief and public GitHub Issues

No real-world usage data or performance metrics are provided.

Evidence: Not evidenced.

The description does not provide any information about competitors or the competitive landscape. It does not mention similar tools, platforms, or methodologies in client discovery or project planning.

There is no discussion of how Manage Client Discovery compares to existing solutions such as:

• Traditional project management tools
• AI-powered interview planning tools
• Knowledge management systems

Evidence: Not evidenced.

1. Unverified claims: All descriptions are self-reported and unverified.
2. No real-world usage: No evidence of adoption, customer feedback, or actual use beyond a demo loop.
3. Limited scope: The tool appears to be designed for a very specific workflow within Codex, with unclear scalability or broader applicability.
4. Single-person team: The project is built by one person (Ayami Sekiguchi), raising questions about long-term maintenance and development capacity.
5. Unclear monetization path: No indication of how the tool will be monetized or scaled.
6. Dependency on GPT-5.6: The system relies heavily on a proprietary AI model, which may not be available in all environments or markets.

Inference: These risks are derived from the lack of evidence for traction, scalability, and business viability — not from any direct observation.

1. How many actual client discovery sessions have been conducted using this tool?
2. What is the current level of user feedback or iteration based on real-world usage?
3. Are there plans to integrate with other tools beyond Codex, and if so, what are they?
4. How does the tool handle edge cases or unexpected inputs during interviews?
5. Can you walk us through a typical workflow from start to finish in a real project setting?
6. What is the long-term vision for this product — will it remain a plugin or evolve into something more scalable?