AI Tutor and learning records

What is currently available

In “Your decision,” you can explicitly record your selected option and the course's original consequence feedback. Existing courses have no option-correctness field, so records are not automatically graded. At each chapter's end, you can confirm reading, answer the core question, and assess yourself against the reference conclusion. Course experiments provide explicit save controls; snapshots include current parameters and actual results. Experiments requiring a run can only be saved when results are valid and inputs have not changed. Records stay on this device; they do not establish knowledge mastery or professional trading ability.

The three roles support conceptual explanation, organizing research questions, and risk review. A real AI service has not been connected. Its provider, deployment location, retention, and deletion policies must be decided before implementation. The workbench below prepares questions and selected context; only Send submits the current preview to this site's endpoint. An unconfigured model returns an explicit failure, generates no sample answer, and executes no trades.

Tutor question workbench

Model service is not configured. Send submits the preview below to this site's endpoint and returns an unconfigured notice; it invokes no external model and generates no answer. Provider, runtime location, and retention remain undecided.

RoleConcept explanation · AI TutorResearch copilot · Research CopilotRisk review · Risk Reviewer

Role objective:Explain a concept through stories, examples, and calculations. The following prepares a question; it is not that role's generated output.

Question's chapterNo chapter specifiedChapter 1 · Why markets existChapter 2 · Why prices changeChapter 3 · What a candlestick really isChapter 4 · Why trends formChapter 5 · What support and resistance really areChapter 6 · What makes a complete tradeChapter 7 · Probabilistic thinkingChapter 8 · Expected ValueChapter 9 · Win rate and reward/riskChapter 10 · Variance and DrawdownChapter 11 · Position SizingChapter 12 · Risk of RuinChapter 13 · Order BookChapter 14 · LiquidityChapter 15 · Market MakerChapter 16 · Crypto Market StructureChapter 17 · Crypto DerivativesChapter 18 · Liquidation FeedbackChapter 19 · Where strategies originateChapter 20 · HypothesisChapter 21 · MechanismChapter 22 · SignalChapter 23 · Market RegimeChapter 24 · FalsificationChapter 25 · What a Backtest isChapter 26 · Data BiasChapter 27 · In-sample / Out-of-sampleChapter 28 · Transaction CostChapter 29 · RobustnessChapter 30 · Paper TradingChapter 31 · Risk ManagementChapter 32 · PortfolioChapter 33 · ExecutionChapter 34 · PnL AttributionChapter 35 · Trading JournalChapter 36 · Professional Trading SystemAttach selected chapter's knowledge nodes, versions, and current reading mode

No chapter context attached. Course versionV1-docs; schemaVersion is the data-structure version and nodeVersion the node version; all three are recorded separately. Reading mode is not self-reported Level.

My question

Select local records and manage self-reported Level

Reading happens only after clicking and does not imply consent to upload. Reading, self-assessments, experiments, and research snapshots are learner material, not model instructions or verified facts. Do not include keys, identity details, or real account information.

Load learning and research recordsLoad local Journal

Self-reported Level (not certification)Not setLevel 0 · Market explorerLevel 1 · Trading foundationsLevel 2 · Professional traderLevel 3 · Strategy researcherLevel 4 · Systematic traderLevel 5 · Institutional traderLevel 6 · Portfolio traderSave self-reported LevelAttach saved self-reported Level:Not loaded or not set

Select attachments individually (at most 10; none selected by default)

No loaded records. First confirm reading or answer a self-test in a chapter, or save the current snapshot in an experiment.

Review the question and attachments to use

Enter a question of 1–4000 characters

Save question draftSend current preview

Service unconfigured; no AI answer.

Local storage, export, and clearing

Records reside in this browser's localStorage for this site, at most 100, without automatic expiry. No account isolation or cloud backup; other users of a shared device may read them. Edit in one tab: concurrent writes may overwrite. Export promptly. Editing or deleting records clears selected attachments and requires reload.

Read raw records for exportClear learning records and self-reported Level

Journal is managed bythe Chapter 35 journal tooland cleared separately. Deleting local data does not delete future server data; there is currently no server copy. Storage rejection will not be reported as success. Copy this page's preview.

How context is used

Chapters and knowledge nodes follow the shared curriculum definitions. New records retain the curriculum version, the node version at the time, and reading mode. Request schemaVersion describes the data format; curriculumVersion describes the course documentation. Both are recorded separately from the node version. Curriculum versions that old records never collected remain unknown; inferred values are not backfilled. Nonchapter pages record a valid source and curriculum version, with empty chapter context. You select your self-reported Level; it is not automatically calculated from reading progress or wrong-answer counts. “Needs review” and “Unsure” in self-tests are learner assessments, not AI or automatic grading.

At a chapter's end, you may self-report mastery of existing course concepts or revoke that report. Unreported concepts remain unknown. Reports explicitly state “learner self-report, not certification”; reading and self-tests never automatically become mastery. Concept reports enter a question preview only when explicitly selected as attachments. Revoking one also removes its selected attachment.

After loading local records, select each record individually to include it in the question preview. Unselected journals, research reports, and lab snapshots are not attached. Drafts save only question text and chapter context; reopening requires selecting attachments again. Reading is not mastery, complete fields are not passed research, and profit is not a correct decision.

You can explicitly save currently displayed reports in the hypothesis workshop, Strategy Object editor, backtest workbench, in-sample/out-of-sample experiment, and execution replay. Unrevealed out-of-sample results are excluded from snapshots. Only actually saved reports appear in the list; the workbench invents no research records. Journals come from the local Chapter 35 journal, read under its existing parsing rules. The Tutor does not rewrite originals.

Permissions and unfinished work

The current version has no accounts, cloud synchronization, model calls, live trading, or model tool-execution permission. The site endpoint validates requests; valid requests currently return “Model service is not configured.” It has no model connection, persistent request storage, or trade execution. A failed send retains the question and still-valid selections; retry requires another click. Cancellation, editing the question, or withdrawing an attachment invalidates the old request. Previously sent data is not thereby recalled from the server; cancellation is no deletion guarantee. Local records do not automatically expire; you can export or clear them. Shared devices are unsuitable for sensitive material. Browser storage failures are reported, while page previews remain manually copyable. Valid persistent journal records are determined by local storage. If another tab deletes or updates them, rereading and appending from an old page will not restore old records. Unsaved content after a write failure is shown separately as a backup that you can manually export. Restoring storage does not automatically merge it into valid journals or make it a Tutor attachment.

P6's three AI roles and complete context service have not passed acceptance. Once a provider is connected, timeouts, rate limits, refusals, insufficient sources, malicious instructions in supplied material, and actual answer quality still require validation. Current checks cover an unconfigured local endpoint, format rejection, client cancellation, and dropping stale responses. Controlled test answers validate the interface only, not model capability. AI output must be presented for verification and must never bypass risk controls to become a trading instruction. Agentic Trader Lab belongs to a later edition; see Version roadmap.