Research goals, literature and strategy

Open Today’s research → Research suggestions in the Reports and actions row → + Paper goal. Select the lab and related projects, research question, venue and year, and your submission target date. Record available data, GPU time, collaboration constraints and PI/researcher perspective under Research time, resources and writing criteria. Goals are private and grant no additional source access. Use the Paper targets tab to set annual targets by category; legacy ambitious and other targets remain unclassified. Targets do not mean acceptance or publication; update the actual stage yourself.

Enter general English topic terms that may be shared publicly and enable Recent literature checks. Only those explicit terms go to OpenAlex, not unpublished research text automatically extracted as a search query. Checks enter a queue on a nominal 30-minute per-goal cadence and retrieve up to 25 results published in the past 14 days. Indexing delays, provider limits and queue load mean neither instant discovery nor exhaustive coverage is guaranteed. Discovered literature shows the last successful check and pending analysis count. Also enable automatic application advice/strategy to compare papers with your evidence and suggest an application, decisive experiment and limitations. A new-paper alert is created only when the analysis identifies concrete application value. Discovery alone uses no AI allowance.

Enter a question such as “Suggest a paper and essential experiments feasible within ten days” and choose Strategy analysis. It uses bounded excerpts from currently accessible stored material, reports, personal notes and action results. With literature checks enabled it also searches earlier relevant work using the same public terms. Supported official venue/publisher HTML pages can supply a submission-guidance excerpt. Unread links and papers without full text remain explicit coverage limitations; registering a URL does not verify the rules. The report examines claims, prior-work differences, counter-evidence, decisive experiments, resources, critical path, fallback scope and potential overlap between papers. Feasible, conditional, rescope and evidence-needed are AI assessments, not acceptance probabilities. Indexes such as KCI and Scopus are not a single research-quality ranking.

Enable literature checks and automatic AI separately for each goal. Background analysis shares your personal allowance and must satisfy both a daily background cap and a reserve for manual use. Defaults are one background analysis and one reserved manual attempt per day. When allowance is unavailable, discovered papers remain stored for later analysis. A new analysis is marked as one attempt; reopening matching saved results needs no additional AI. Failed started attempts follow the existing quota policy. Quiet hours use your Today’s research notification timezone. Daily digests combine ready suggestions after the selected hour; later arrivals wait until the next day. Email also requires existing email preferences and a configured sender; push requires a registered device and enabled notification category. Archiving a goal stops future discovery/automatic analysis while retaining its history.

Open Evidence and application from Research suggestions to inspect numbered citations and the source list. Act under Next actions opens the existing action editor; review and approve the destination and content before an external comment or message is posted. Add to today’s plan appends the summary and actions without replacing existing notes. Adopt marks personal intent to use a suggestion, not paper acceptance or experimental completion. Use Already in progress, Show next week or Not interested to manage suggestions. Saved proposals inform AI consultation and plan/review drafts while remaining distinct from confirmed results. Subsequent execution results and new research records are reconsidered during later strategy analysis.

Choose daily, every three days, weekly or fortnightly reviews. When automatic analysis is enabled, targets within seven days of the deadline are considered daily; changed internal evidence is reconsidered at least a day after the preceding analysis. New-paper analyses share the same personal cap. Earlier reports remain in the goal’s version history. The ChatGPT connector can read saved goals/reports and prepare goal or personal-setting changes and a strategy analysis. Commit only after approving the exact review and cost. Poll the same goal to read analysis status instead of creating another request. External publication still needs the existing separate action preview and approval. If tools are unavailable, check reconnection and granted scopes.

If a successful check finds nothing, broaden the public topic terms. Provider errors retain the last successful timestamp and retry on the next check. Papers waiting for analysis require automatic analysis enabled, sufficient cap/reserve and a working AI connection. Failed analysis retains the underlying evidence; correct the conditions and request analysis again. Changing project scope or source permissions may hide earlier reports; analyze the current scope again. Background discovery and analysis require the server worker to be running. The web page need not remain open.

In Paper targets, select a year and View, then Add target. Choose KIISE conferences, SCI(E)/SCIE, Scopus, KCI or Custom and enter a submission count. Details hold the venue, official tier and edition, Q1 field, KCI candidate inclusion, main conference/workshop and reference link. These are user-defined criteria, not automatically verified accreditation. Keep one goal per manuscript and choose one primary annual counting category. Edit that same goal for resubmission, retaining its first submission date. Enter only confirmed submission, acceptance and publication dates; undated legacy stages are excluded from counts. Changing years preserves earlier settings and dates. Removing a target row removes only that annual target, not research records.

Under AI allowance, set the daily automatic cap and reserve for direct requests. With a total cap of 50, zero used, automatic cap five and reserve five, up to five automatic attempts are available. After 43 direct attempts, only two of the remaining seven can be automatic, leaving five. A total cap of five with a reserve of five stops automation. The reserve is not extra allowance and does not cap manual use. Automatic cap zero disables automation; reserve zero protects no allowance. Live calculations use the current usage snapshot and can change if you use AI elsewhere. The worker rechecks current quota and permissions.

Click or tap ? beside a setting for its meaning and examples. Keyboard Enter opens it and Escape closes it. In Notifications, choose delivery after analysis or a daily digest. Digest time uses your personal time zone and actual delivery follows the worker schedule. Quiet hours take precedence; equal start and end times disable the quiet interval. This does not pause analysis. Email also requires personal email preferences and a platform sender; it uses no extra AI. Save paper targets separately from AI/notification settings. Concurrent changes require a reload instead of silently overwriting saved settings.

New settings start with one target row. Add rows as needed; check Exclude this target and save to remove only the category. Uncheck before saving to undo. Existing targets are preserved. In Record progress, select the manuscript and enter owner-confirmed first submission, acceptance and publication dates. Each event counts once in its calendar year; a stage without a date is not counted. No external submission system automatically verifies these events. Reuse the same goal for resubmissions. Trashing a paper goal retains linked reports/code; restoration keeps automation off.

Only platform administrators control the shared public-analysis daily limit. Shared analysis of public papers/code has service cost but does not deduct personal AI attempts. Personalized strategy/code analysis uses personal quota. Expand Detailed report and downloads to set a maximum attempt budget for four saved report parts. Resume from saved parts after a budget stop or failure. Calls already started may still be charged after cancellation. Length and citation count follow available evidence; neither is padded with invented sources. ZIP includes Markdown, HTML, PDF, source JSON and BibTeX. Output examples are explicitly synthetic, not measured research, and cost no AI.

Pages show saved results first and poll a small status response about every eight seconds while active or thirty seconds while idle, pausing in hidden tabs. Open View new results to refresh saved output. Polling uses no AI. ChatGPT uses guidance from consultant_strategy_list, status/dossier reads, and the same prepare/commit approval contract for progress, trash/restore and dossier start/pause/resume. If the host does not expose these tools, refresh connection metadata; server support alone does not prove host availability. Download links require a current platform login.

Select grants/projects to include their purpose and constraints. The question editor supports mentions, document citations, Markdown tables and image/file attachments. Example: “Compare the model on page 6 with our project.” After saving, expand Research context to review the scope. Text extraction runs in the background after upload; analysis waits for pending inputs. Failed inputs can be retried. PDF text has extraction limits; the requested page or one question-relevant page is visually inspected. Check page numbers and omissions. Purpose, ideas, self-reports and validated results remain distinct; mentions do not send messages. Identify priority files/pages for large collections.

Staged strategy analysis saves the original question before running four parts: literature comparison; question/contributions/scope; model/experiments/implementation; schedule/integrated strategy. The default maximum is four personal AI attempts, one per stage. A lower cap pauses after saving completed parts. Waiting for another job, attachments or personal allowance does not lose your question. The compact status shows the target, current stage and saved parts. Expand request history for manual/automatic origin, original question, timestamps and errors. Attachment preparation and analysis completion are counted separately. Intermediate stages save concise, focused parts before final integration. Each stage has an AI response limit of up to eight minutes and continues after closing the page; this is not an estimated completion time. Timeouts pause; retry the saved request or resume from saved parts. Changed conditions require a new request. Download partial results as ZIP or Continue discussion without overwriting the original. Reading costs no platform AI; new analysis and consultation use personal allowance. Automatic failures back off and stop after three consecutive failures for the same goal revision. Daily AI review considers accessible, date-bounded goals and annual-target changes, analysis attempts/results, code proposals, working revisions and decisions. Missing historical snapshots or bounded coverage are not represented as complete collection.

Use Share and collaborate on a strategy report to review content, recipients and read or follow-up permissions. Report ID and immutable edition ID/version preserve lineage. Original attachments, private chats and GitHub access are not automatically shared. Open Shared research reports, then start your own discussion or code workspace. Share a saved result with the team by selecting an exact revision/checkpoint/code result and files to return to the author. Seen, review and adoption are separate human actions. Sharing, reading and exporting make no AI call; new generation uses the initiating user’s allowance. Stable request IDs prevent duplicates. Revocation, expiry, withdrawal and project access changes affect subsequent access and continuation. Exported copies cannot be recalled. Notion queues a fixed edition under an authorized parent; unknown outcomes never auto-retry. Reconcile the existing page by ID. External researchers must first complete existing account and lab/project participation approval; member rollout is not automatically enabled.

The live strip shows queued/running requests only. Analysis archives separate attention, completed, cancelled/failed and trash; saved outputs remain downloadable. Cancel before moving active work to trash; restore never resumes automatically. Permanent deletion requires an exact scope review and current permission, and rejects references from later consultations/code. Original attachments and usage totals remain. Old encrypted backups are not immediately erased; reapply the latest deletion ledger after restoration. ChatGPT uses consultant_strategy_prepare analysis_trash/analysis_restore/analysis_purge and commit.

Goal settings precede research context. Paginate all literature/search observations and reports. Deduplicated literature and per-search snapshots are separate; legacy observations have unknown original search batches. Only the final analysis stage creates a final report; partial results remain on the request. Compare two saved reports without AI. ChatGPT strategy_read collection=reports/literature/observations with goal,page reads the same lists.

New strategy analyses default to prior reports, feedback, saved follow-up decisions and recent discussions for the same goal. Originals remain available; bounded excerpts and omissions are recorded for model input. Independent analysis excludes prior analysis context. Revoked evidence is never reused. Code consultation from a report, section or action retains source provenance. ChatGPT resumes the same request ID and checkpoints; completed work uses follow-up analysis. analysis with status=true returns small metadata. Review strategy_prepare acknowledge with kind/id/version before commit. If the host exposes an old limit or missing operations, refresh the connected tools and read the central policy.

Each code workspace has one representative notice across preparation, AI and validation steps. Dismiss preserves failures and does not mark them successful. Link a strategy report under Sources and links; the report also links back to derived code. External results preserve the question, answer, files, provenance, versions and downloads. Ask ChatGPT or Codex to continue this LUNA workspace and save the answer and files to the same workspace. Only a ready storage receipt confirms synchronization. Conversations or temporary files never delivered to LUNA are not automatically imported. You can also paste an answer and select files in the platform. Conversation text is limited to 120 KB, files to 100 files and 8 MB total; retry with identical files and request identity to resume. Files are not executed; model and test claims are producer-reported. Revisions preserve previous results. Storage uses no new AI call but consumes storage and backup capacity. Encrypted rejected responses may be revalidated without AI for up to 24 hours; invalid responses remain failures. Execution/resume settings start collapsed.

Analysis detail shows unavailable evidence counts and Google reauthorization status. Open the linked connection, personally authorize Google, wait for collection, then check access again. Authorization alone never restores historical access. Reuse saved stages only after a new authorized collection confirms the same file, version and content. Changed content or revoked access keeps old results preserved but restricted. Review a new analysis with accessible evidence to create a separate four-stage request after reviewing the original question, budget and timeout. Blocked source text and old outputs are not copied; new calls consume the existing personal allowance. Preview/status reads never run AI. ChatGPT uses strategy_read analysis diagnosis and the exact strategy_prepare analysis_recover / strategy_commit review contract.

Use the automation control at the top of a paper goal without deleting it. Conditions are collapsed by default. Periodic review offers None, daily, every three days, weekly and fortnightly. None disables periodic review only; new literature, changed evidence and deadlines within seven days are independent triggers. Disable automation to stop every automatic trigger. Literature discovery is separate and does not consume personal AI attempts. Disabling cancels unstarted automatic requests; a started call may incur its existing cost and save a validated result before stopping. Manual analyses, reports, literature and history remain. Enabling or changing conditions schedules future work without replaying missed periods. Allowance waits, failure holds and running states are distinguished; the next review time is an estimate. Ask ChatGPT or AI consultation to disable automation for the goal. consultant_strategy_read returns automation and its revision; prepare(operation=goal, values={automation_revision, automation:{automatic:false}}), review the exact change, then commit updates the same policy. Actual host tool exposure must be verified separately. Reading and saving policy do not execute an AI analysis.

Editing goal settings preserves discovered papers, analysis links and observations. Titles and deadlines do not change the search revision; changing search terms does. Accumulated literature is the default, with filters for current, earlier or unknown settings, recent discovery, analysis availability and relevance review, paginated by 20. Unknown historical search terms are never inferred. Prioritize, select by relevance and exclude apply to future analyses without deleting previous results. Reassess relevance after changing the question or research scope. Reads and selection saves spend no AI allowance; automatic analysis consent is separate. consultant_strategy_read(goal) returns the same library in ChatGPT and AI consultation. prepare(operation=goal, values={literature:{operation:list,filter:all,page:1}}) is read-only. For changes, use literature operation=select, paper, inclusion(auto/include/exclude) and revision, review, then commit. Search observations retain the actual terms and settings.

Use the top share icon for team sharing or Notion publication. Find recipients by name and use ? for field help. For Notion, select an allowed parent and review the body before queuing publication. Configure a connection if none is available. Notion audiences are independent. The export icon provides full-body print, PDF and Markdown. An uncertain publication is reconciled, never blindly reposted.

Connected tools support reading a goal workflow and recording notes or imported results. Use the lifecycle envelope inside consultant_thread_prepare save, followed by reviewed commit for changes. Only transmitted content is saved. Reads and record writes do not invoke extra AI or remote experiments; new consultation follows existing usage policy.