Manage analysis progress, pause/resume and daily allowances.
Reviewed · 2026-10-12
On this page · 18
Choose analysis and inspect progress
- Platform administrators open Lab management → Material AI analysis. Select metadata, content or both and expand categories to check the intended scope.
- Metadata analysis creates proposals for dates, categories and tags; approval is required. Content analysis creates notes about text, pages and images.
- Distinguish completed, running, queued and failed items. Inspect status, errors and source dates. Preserved historical results excluded from current materials after collection failures are marked separately.
- Pause/resume affects that analysis track. An in-flight response may finish; resuming does not regenerate every completed result.
Normal allowances and the fixed initial cohort
- Check the current shared analysis and personal AI allowances. Normal content analysis uses attempts per page, image or text segment, including failed attempts where applicable.
- A fixed initial cohort has a separate daily budget for both metadata and content. The default and maximum is 100 materials; zero pauses it. Multiple pages of one material count as one material that day.
- New materials and newly observed versions do not inherit this exemption. Material counts are not exact token/cost limits or guaranteed completion counts.
- Quota-bound work resumes through the worker after the Seoul date changes. Permissions, source changes, configuration and pause controls still apply.
Separate registration, consultation and backup
- AI analysis of direct uploads is optional and off by default. The uploader or a platform administrator requests it; normal items use the uploader's allowance.
- Notes can be reused as consultation evidence candidates, but every file is not read in full for each question. Inspect omissions and partial coverage.
- Check preservation and Drive-copy status separately from completed analysis. Shared research copies and private consultation storage have different boundaries.
- Analysis is AI interpretation. Compare source dates, figures and numbers and approve only appropriate metadata proposals.
Personal AI and shared analysis allowance policy
- Settings → AI usage and remaining allowances is available to every user for personal usage, remaining attempts and their labs’ shared analysis budgets. Platform administrators manage personal allowances.
- AI help, personal AI consultations, announcement/document translation, reports and magazines use personal allowances. Selecting a lab in AI help chooses its execution configuration and does not spend its analysis budget. The legacy separate 10-attempt cap no longer applies.
- Material AI analysis → Daily analysis attempt limit for regular materials is the sole source of the shared analysis budget. Saving 500 means 500 attempts per day across metadata, file, text and automatic conversation analysis. Existing personal allowances also apply to regular analysis; direct uploads are charged to the uploader. Regular analysis therefore also waits if the personal allowance runs out first.
- A fixed initial cohort uses a separate limit of up to 100 materials per day, outside personal and regular analysis allowances. New materials and versions do not inherit that exemption.
- Daily usage resets at midnight Seoul time. Failed execution attempts count, but permission or allowance rejection does not. Queued and duplicate requests are counted only when an execution starts. Long translations and multipage analysis may need multiple executions. On-device browser translation and handbook search use no AI attempts.
- Earlier failures remain historical records. Ask again if personal attempts are currently available. Provider subscription/rate limits are separate from app allowances; AI permission and an enabled lab configuration remain required.
AI management: settings in one place
Personal settings → Personal AI settings always shows only your effective policy, translation and usage, even for administrators. Lab management → Platform AI management is a separate administrator page for execution/resume policy for all users, platform statistics, user permissions and prices. Edit notifications on the dedicated notification settings page. A request budget is separate from the daily allowance. New work snapshots saved defaults; only platform administrators can change the central policy. Administrator defaults are 50 attempts and 3000 seconds per call, with no overall elapsed-time deadline. Deliberate resume applies current policy without resetting usage. Reductions cap subsequent calls of existing requests. Completed stages are reused.
Attempts, tokens and estimated costs
Attempts are not invoices. New executions record feature, user, grant, project and requested model, with daily/weekly/monthly/yearly or custom-date views. Filtered totals above and below the table cover calls, input/output tokens and estimated USD across the complete current filter, not just displayed rows. Check measured-call coverage against all calls for each metric. Unknown tokens and costs remain unmeasured; historical gaps are not zero. Administrator-entered rates with a source/date apply to estimates for subsequent executions. Subscription fees and invoices are separate.
Paused work, source access and attachments
Analysis detail preserves the request and work status. If source access is unavailable, affected output is hidden with connection/permission guidance; it is not deleted. Recheck after access is restored. Request history distinguishes loading, empty, failed and retry states. PDF page failures preserve successful page extraction and show causes such as encryption, timeout or pages without text. Extracted text, visual inspection and experimental validation are distinct.
Personal settings through ChatGPT and AI consultation
Ask “Show my remaining AI allowance and reset time”, “Show my contribution score”, or “Show my administrator-managed execution policy”. Shared account tools read current settings and prepare an exact before/change review. Execution policy changes require current platform administrator authority and explicitly review the effect on all users. Personal overrides are rejected. Other settings remain personal; sharing rights and publication remain separate. Tools: consultant_account_read, prepare, commit, usage, contributions and diagnose. Help AI can read current personal settings and usage. Continue to work/settings consultation carries the question into private full consultation; subsequent reasoning uses personal AI allowance and changes require review. If the ChatGPT host does not expose a tool, check the connection and granted scopes.
Repository consultation, isolated checks and integration
In GitHub code consultation choose architecture/paper mapping, README, function/class documentation or commit/research changes. The inspected file coverage and up to 20 commit messages are explicit; execution is never inferred. ZIPs include inspected source plus changes under standalone/, INTEGRATION.md, patches and check receipts. Run toy examples and tests against the real generated API in an isolated environment before integrating on a separate branch at the pinned baseline. Zero tests or missing dependencies do not establish verification.
Detailed filters and research notifications
Detailed filters search authorized users, grants and projects even without recorded usage. Select all translation or individual features including help. Unassigned means missing attribution; unmeasured tokens are not zero. The top area shows active work, issues and unseen results. Mark as seen acknowledges a notification, not scientific approval. Cancelled and reviewed work remains in the goal/code history and archive. Saved materials are preserved.
Analysis continuity and ChatGPT
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.
Synchronizing external discussion and code results
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.
Calendar ranges and official reference prices
Today/this week/this month/this year set date bounds in Asia/Seoul. Choose chart granularity separately in advanced filters. Administrators receive OpenAI Standard reference defaults checked on 2026-10-05. Custom rates and historical estimates are preserved. Long context, cache writes and processing tiers are distinguished; unknown tiers use an API-equivalent estimate, not subscription billing.
Active work, archives and deletion
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.
Complete literature and report history
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.
Start repository consultation
Choose a project, repository and purpose using Start GitHub repository consultation on the code landing page. Continue an existing analysis where available. Consultation is expanded on the detail page and requires a completed repository read. Connection/permission problems link to connection settings. The current bound of 24 text files/85KB does not establish whole-repository verification.
Evidence access and analysis recovery
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.
Research daily limits and the overall allowance
Use Manage limits in Today's research to open Daily usage limits in Platform AI management.
Administrators choose Follow overall AI allowance or Set a separate cap. Administrators inherit by default; the existing regular-user default remains 5. Inheritance is not extra credit: research shares the personal pool with translation, help and other AI. A separate cap of 0 disables research AI.
Manage my overall daily AI allowance opens per-user overall limits. Attempted runs, queued reservations, completed and failed runs and reset times are distinct.
The overall pool resets at Seoul midnight; a separate research cap resets at midnight in the research timezone. Per-request budgets and per-call timeouts remain separate execution settings. Raising limits never clears history or usage, or automatically retries failures.
In ChatGPT or AI consultation, read research_usage via consultant_account_read. Only administrators can review and commit execution changes to research_daily_limit/admin_research_daily_limit; null inherits the overall pool.