"Afraid to Ask" designed without Analyze

specs/004-afraid-to-ask/spec.md

@@ -0,0 +1,86 @@
+# Feature Specification: Afraid to Ask
+
+**Feature Branch**: `004-afraid-to-ask`  
+**Created**: 2025-10-13  
+**Status**: Draft  
+**Input**: User description: "Afraid to Ask. You can find requirements here: .context/afraid-to-ask.md"
+
+## User Scenarios & Testing *(mandatory)*
+
+### User Story 1 - Submitting an "Afraid to Ask" Idea (Priority: P1)
+
+A user wants to express a sensitive idea or desire without it being widely visible to others unless there is a clear alignment with other users' expressed desires. This allows them to gauge interest or acceptance for potentially sensitive topics in a private manner initially.
+
+**Why this priority**: This is the core functionality of the feature, enabling users to input their "Afraid to Ask" ideas. Without this, the feature cannot exist.
+
+**Independent Test**: A user can navigate to the response form, input an idea into the "Afraid to ask" field, and submit the form. The system should record this idea with the specified privacy settings, and it should not immediately appear in public results unless compliance conditions are met.
+
+**Acceptance Scenarios**:
+
+1.  **Given** a user is on the response form, **When** they enter an idea into the "Afraid to ask" field and submit the form, **Then** the idea is successfully recorded by the system with maximum privacy.
+2.  **Given** a user has submitted an "Afraid to ask" idea, **When** they view the results before any other user's "Want" or "Accept" ideas comply with it, **Then** their "Afraid to ask" idea is not publicly displayed in the results.
+
+---
+
+### User Story 2 - Viewing Results with "Afraid to Ask" Ideas (Priority: P1)
+
+Users want to see if their sensitive ideas align with others' desires, or if others' sensitive ideas align with their desires, without explicitly revealing the "Afraid to Ask" nature of the idea until a match is found. This promotes discovery of shared, sensitive interests.
+
+**Why this priority**: This story defines how the "Afraid to Ask" ideas become visible and useful, providing the value proposition for the feature.
+
+**Independent Test**: Multiple users can submit ideas, including "Afraid to ask" ideas and "Want"/"Accept" ideas. The system can then be queried to display results, and the visibility of "Afraid to ask" ideas should correctly reflect the compliance logic.
+
+**Acceptance Scenarios**:
+
+1.  **Given** User A has submitted an "Afraid to ask" idea, and User B has submitted a "Want" or "Accept" idea that complies with User A's "Afraid to ask" idea, **When** any user views the results, **Then** User A's "Afraid to ask" idea is displayed as if it were a "Want" idea from User A.
+2.  **Given** User A has submitted an "Afraid to ask" idea, and no other user's "Want" or "Accept" idea complies with User A's "Afraid to ask" idea, **When** any user views the results, **Then** User A's "Afraid to ask" idea is NOT displayed in the public results.
+
+---
+
+### Edge Cases
+
+-   What happens when a user submits an empty "Afraid to ask" field? (It should be ignored or validated as empty).
+-   How does the system handle multiple "Afraid to ask" ideas from the same user? (Each should be evaluated independently).
+-   What if an "Afraid to ask" idea complies with multiple "Want" or "Accept" ideas from different users? (It should still be treated as a "Want" and displayed).
+
+## Requirements *(mandatory)*
+
+### Functional Requirements
+
+-   **FR-001**: System MUST add a new input field labeled "Afraid to ask" to the response form, positioned directly below the "What you want" field.
+-   **FR-002**: System MUST store "Afraid to ask" ideas with maximum privacy, meaning they are only visible to the submitting user until they "comply" with another user's "Want" or "Accept" idea.
+-   **FR-003**: System MUST compare "Afraid to ask" ideas from one user with "Want" and "Accept" ideas from other users using semantic similarity (e.g., via an LLM to compare meaning).
+-   **FR-004**: If an "Afraid to ask" idea complies with any other user's "Want" or "Accept" idea, the system MUST treat and display it in the results as if it were a "Want" idea from the submitting user.
+-   **FR-005**: If an "Afraid to ask" idea does NOT comply with any other user's "Want" or "Accept" idea, the system MUST NOT display it in the public results.
+
+### Key Entities *(include if feature involves data)*
+
+-   **Idea**: Represents a user's input, can be "Want", "Accept", or "Afraid to ask".
+    *   Attributes: `content` (text), `type` (enum: Want, Accept, AfraidToAsk), `userId`, `privacyStatus` (e.g., private, public).
+-   **User**: Represents an individual interacting with the system.
+    *   Attributes: `id`, `name`.
+-   **ResponseForm**: The interface where users submit their ideas.
+
+## Dependencies and Assumptions *(optional)*
+
+-   **Assumption**: The system has an existing mechanism for users to submit "Want" and "Accept" ideas.
+-   **Assumption**: The system has an existing mechanism to display results based on user ideas.
+-   **Assumption**: "Maximum privacy" for "Afraid to ask" ideas is achieved through ephemeral server-side storage, meaning data is encrypted, stored temporarily on the server, and purged on session termination, making it inaccessible via client-side developer tools.
+-   **Dependency**: An LLM or similar semantic comparison service is available for evaluating idea compliance.
+
+## Clarifications
+
+### Session 2025-10-13
+
+- Q: What constitutes "complying with" when comparing "Afraid to ask" ideas with "Want" and "Accept" ideas? → A: Semantic similarity (e.g., using an LLM to compare meaning)
+- Q: How is "maximum privacy" achieved for storing "Afraid to ask" ideas? → A: Only visible to the submitting user until it "complies"
+
+## Success Criteria *(mandatory)*
+
+### Measurable Outcomes
+
+-   **SC-001**: The "Afraid to ask" field is present and correctly positioned on the response form, as verified by UI inspection.
+-   **SC-002**: "Afraid to ask" ideas that comply with other users' desires are correctly displayed in the results as "Want" ideas, with 100% accuracy in testing scenarios.
+-   **SC-003**: "Afraid to ask" ideas that do not comply with other users' desires are not displayed in the public results, with 100% accuracy in testing scenarios.
+-   **SC-004**: The privacy requirements for "Afraid to ask" ideas are met, as confirmed by a security review.
+-   **SC-005**: Users can submit "Afraid to ask" ideas through the response form without encountering any system errors.