Eleven tools that check your work

Eleven capabilities. Every one returns critique, retrieval or structure. None returns text you could submit as your own — and that is a constraint built into the systems, not a policy we ask you to trust.

Why the limit exists

The same technology that critiques a proposal can write one; the difference is a single instruction. If the restriction lived only in our terms of service it would be circumvented within weeks, and we would have built exactly the service our commitment page says we refuse to offer.

So it is implemented at the system layer: each capability is scoped to an output type, and those scopes are tested automatically before release. A capability that cannot be safely constrained does not get built.

The eleven capabilities

Title Workshop S2	Critique of your own draft title against discipline conventions. Does not propose titles.
Gap Validator S3	Tests a gap you have stated for evidence, significance and feasibility. Does not identify gaps for you.
Literature Navigator S3	Searches, retrieves, organises and summarises real sources. Cites only retrieved sources. Never generates a reference. Refuses to draft review prose.
Design Selector S6	A decision path from your research question to an appropriate design. Returns structure and rationale, not written methodology.
Hypothesis Checker S6	Checks direction, testability, falsifiability, variable alignment and matching test. Does not propose hypotheses.
Questionnaire Checker S8	Item-level diagnostics — and first, a search for a validated instrument. Flags problems and suggests item types. Never returns finished items.
Statistical Advisor S10	Test selection, assumption checks, and interpretation of your output. Does not write the results narrative.
APA Assistant S11	Format correction of references you supply. Mechanical formatting only.
Proposal Structurer S4	Section scaffold, prompts and per-university requirements. You write every word of content.
Theory Explorer S3	A browsable map of theories by discipline and construct, with citations. Presents. Does not recommend or select.
Research Roadmap Generator S1	A lifecycle plan with dates and artefacts for your stage. Planning output only.

The AI Usage Statement

Every interaction is logged with the capability, timestamp, input and output. You can export that as a dated record and attach it to a submission or show it to your guide.

Universities are moving towards requiring declaration of AI assistance. Most tools make that declaration difficult. This makes it a document you can produce on request — which is only possible for an operator with nothing to hide.

Grounding

• Everything is grounded in verified content and cited sources. No ungrounded generation on procedural questions.
• Where we have not verified an answer, the output says so and tells you how to confirm with your university, rather than producing a plausible guess. Accuracy on procedure is the one thing a platform like this has to get right.
• Institution-specific answers resolve to a cited ordinance with its date, or decline.

Availability

These are released in stages, each gated on accuracy testing rather than a date. Until a capability is live its page says so. The free calculators work today and always will.