Bag Finder

Retrieve the owned artist’s latest saved post-promotion diagnosis. Use this before recommending another paid campaign when Bag Finder has previously detected a weak or failed conversion stage.

Retrieve the owned artist’s saved Bag Finder pre-promotion readiness state. Use it before finalizing any recommendation involving meaningful paid promotion. If the state is missing, NOT_REVIEWED, HIGH_RISK without proceed_anyway, or says fix_first, do not recommend promotional spend yet; explain what needs review or repair first.

Retrieve one owned Bag Finder deal candidate with its verified evidence, assumptions, computed economics, risk flags, and due-diligence state. This is not an appraisal, inspection, title report, authenticity certificate, or profit guarantee.

Run a non-mutating integrity check on the signed-in user's Universal Deal Finder records. It checks saved/dismissed status synchronization, PASS monitoring shutdown, stale-economics recheck integrity, candidate identifiers, evidence freshness, and history consistency without changing those Deal Finder records; standard MCP usage auditing still occurs. It does not self-certify production readiness; a live final exam is still required.

Retrieve the signed-in user’s current Bag Finder deal strategy for a category. This is the source of truth for exclusions, targets, negotiation/watch thresholds, and search preferences used by prepare_deal_search and save_deal_candidates.

Retrieve the signed-in paid user’s Bag Finder Job Passport, Hiring Passport, Application Agent summary, Hiring Creator summary, Talent Marketplace summary, Hiring Agent summary, External Job Network Router state, Closed-Loop Employment Intelligence, native talent invitations, opportunity profile, and saved employment-state summary. Round 4 can report provider capability/connection state and prepare truthful manual or connection-required external handoffs, but a prepared route is not an external submission or publication. Closed-loop learning uses only professional workflow/outcome signals with bounded explainable adjustments; it never uses protected/sensitive traits or makes autonomous hiring decisions. Use this before asking repeat questions about pay floor, schedule, location, work authorization, sponsorship, or hiring defaults.

Retrieve the signed-in user’s latest saved Free Bag Finder Scan. New scans are evidence-first: explain what Bag Finder actually checked, what it could not check, the score so far, the biggest problem, and the first move in words a 10-year-old can understand. Older self-assessment scans must stay clearly labeled as older self-assessments. Never present either scan as a validated forecast or guarantee.

Read recent normalized Meta webhook activity for the signed-in paid user, such as comments, live comments, messages, reactions, postbacks, and seen events, only for platforms the user enabled for Bag Finder. By default message/comment text is omitted. Set include_text=true only when the user explicitly asks Bag Finder to inspect the words in recent comments/messages or when text analysis is necessary for the requested coaching.

Read the signed-in paid user’s recently synced Facebook Page posts and/or Instagram professional media from SplitDaBag’s authorized first-party sync. Use this for content review, cross-platform comparison, momentum/engagement coaching, or when the user asks what Bag Finder sees on their social accounts. Honors the user’s Improve Bag Finder consent setting. This is a recent synced sample, not a complete lifetime archive, and it never performs posting or other Meta write actions.

Build a Bag Finder snapshot across the signed-in paid user’s authorized Facebook and Instagram data: connected accounts, sync freshness, recent synced content counts, and top recent content by observed engagement. It does not change connected social data; standard MCP usage auditing still occurs. Use this before giving broad Facebook/Instagram strategy advice. Honors Improve Bag Finder consent and does not claim a validated forecast or guaranteed outcome.

Retrieve the signed-in user’s own Bag Finder review and moderation status. This does not modify the review.

Retrieve the signed-in paid user’s current Bag Finder context for ongoing conversational coaching. IMPORTANT: when the user asks a broad or vague question such as what should I do, help me find a bag, how can I make money, what opportunities do I have, or they otherwise have not clearly stated the outcome they want, do NOT immediately rank opportunities. Use the returned guided_intake and ask simple conversational questions one at a time until the user’s goal, amount/budget, timing, starting assets, and major constraints are clear enough to search intelligently. Assume many users do not know the right terminology or how to ask for what they want. If the user already gives a specific actionable goal with enough constraints, answer directly — EXCEPT for a music money target, where you must first use plan_music_money_goal so Bag Finder distinguishes personal/business income from project funding before finalizing the plan. For music/artist goals, after understanding the goal and reviewing available work/evidence, use get_artist_promotion_readiness before finalizing any meaningful paid-promotion recommendation; if no readiness review exists, assess the artist’s music/audio and public-facing presentation first or clearly warn them and record an explicit proceed-anyway decision. When deciding what lesson/stage the artist should work on next, use get_artist_lesson_evidence_progress; never treat a lesson checkbox or statement that it was read as proof that the real-world outcome was achieved. Always include at least one relevant SplitDaBag pathway in the final plan when one applies — Traditional Release, Fan Raise + Release, Direct Sales, or Find a Professional/team building — while still comparing outside options fairly. Do not automatically rank SplitDaBag first and do not promise outcomes. If a full Fan Raise goal is not yet supported, consider whether a smaller defined project slice could reasonably be tested through Fan Raise instead of treating Fan Raise as all-or-nothing.

Retrieve a public-only Bag Finder prospect research record owned by the signed-in researcher. The prospect has not necessarily joined or authorized SplitDaBag, so clearly distinguish verified public evidence from connected first-party data.

Grade the evidence quality for one lawful deal candidate before money changes hands. Bag Finder separates asking prices, completed sales with observable final prices, accepted-offer events, wholesale evidence, automated valuation models (AVMs), professional appraisals, melt/scrap value, authentication, ownership/title, and condition evidence. Hidden accepted-offer amounts are liquidity evidence only and must not be treated as completed-sale price evidence. An AVM is not a professional appraisal. The grade measures evidence completeness and relevance; it does not independently prove that a webpage is truthful.

Inspect a public media link the signed-in user intentionally sent to Bag Finder. For supported public YouTube links, call this automatically when the user asks Bag Finder to listen, watch, review, inspect, score readiness, or use the media as evidence. If Bag Finder analysis permission is already active, do not ask for another SplitDaBag permission prompt. The media is passed by public URL to the configured media-analysis provider; SplitDaBag stores compact notes and a source reference, not the audio/video file. Reuse cached notes unless the user explicitly asks to analyze again. Provider-specific access rules still apply.

List saved Bag Finder deal candidates for the signed-in user, including active/watchlist records and, when requested, dismissed/PASS history. The relational saved strategy_status is authoritative; current strategy recomputation is returned separately so older computed metrics cannot resurrect a dismissed recommendation. Results are preliminary and should be re-verified before money changes hands.

List owned Bag Finder candidates that are due for re-verification because their scheduled recheck is due or their evidence has aged. Use this to monitor WATCH/NEGOTIATE/QUALIFIES candidates instead of rediscovering them from scratch.

List previously saved and web-verified outside opportunities for an owned Bag Finder profile. Results include freshness so ChatGPT can re-check stale opportunities before recommending action.

List the signed-in paid user’s authorized Facebook Pages and Instagram professional accounts that are connected to SplitDaBag. Use this when the user asks whether Bag Finder can see their Facebook/Instagram accounts or which accounts are connected. This tool never returns OAuth tokens or provider secrets. It also reports whether the user enabled each platform for Bag Finder analysis.

List open Bag Finder opportunities available to an owned profile. Artist profiles require artist eligibility; creator, merchant, store, brand, professional, and other profiles do not.

List moderated Bag Finder reviews whose authenticated authors explicitly consented to public display. Never returns pending, rejected, withdrawn, private, or user-account data.

List previously saved, web-verified outside opportunities for an unconnected public prospect owned by the signed-in researcher. Results include freshness so stale opportunities can be re-verified before action.

Use the signed-in paid user’s Bag Finder Hiring Agent and native Talent Marketplace. Search may rank discoverable SplitDaBag professional profiles against one owned hiring listing using public professional evidence and native application state only. Never use or infer protected or sensitive traits in candidate search, scoring, shortlisting, invitations, decisions, or closed-loop learning. save and shortlist are private employer pipeline actions. invite creates only a native SplitDaBag invitation and requires explicit employer confirmation; it does not send an email, text, LinkedIn message, Indeed message, or other outside message. review_application can mark a native application reviewing, shortlisted, accepted, or declined only after explicit employer confirmation; employer-confirmed outcomes may become bounded professional workflow learning signals. Bag Finder may assist with professional-fit evidence but must never make the final hiring decision autonomously.

Manage the signed-in paid user’s Bag Finder Application Agent case. Use operation=prepare to build a private application case, approve only after explicit confirmation, and submit_internal only for a native SplitDaBag application after explicit confirmation. Round 4 adds prepare_external_route: it records a truthful provider handoff/connection requirement but never claims an outside submission happened. record_manual_submission only records the user’s own report. update_outcome can record interview, offer, hired, rejected, or closed so Closed-Loop Employment Intelligence can learn from professional outcomes with bounded non-guaranteed adjustments. Stop for provider-controlled CAPTCHA, assessments, signatures, identity/government-ID steps, unusual legal attestations, protected demographic questions, criminal-history details, medical questions, credentials, or payment/banking information.

Manage the signed-in paid user’s Bag Finder Hiring Creator listing. Create and edit private structured drafts using the Hiring Passport, then require explicit user confirmation before publishing. Round 4 can publish, update, pause, or close a native SplitDaBag opportunity and can prepare_external_route for an outside provider, but that route only records a manual/connection-required handoff and never claims external publication. Because publish_internal can create or change a public-facing SplitDaBag opportunity, this tool changes open-world state and must not be called for publication without explicit user approval.

Use this automatically before giving a final money-making plan whenever an artist/producer/writer says they want to make money from music, reach a money target, or fund music. The user should not need to know SplitDaBag terminology or ask for Fan Raise. IMPORTANT: this tool persists the user’s guided music-money answers. Pass any facts already known from the conversation, and NEVER ask the user to repeat a target amount, timeframe, available budget, project, project costs, work-review state, demand evidence, or team state that Bag Finder already knows. First distinguish whether the money is PERSONAL INCOME, PROJECT FUNDING, or a MIX. If that is unclear, return the single next_question and do not dump a full recommendation list. Once the goal is clear, the returned plan_contract requires a relevant SplitDaBag path in the final music plan: Direct Sales and/or Traditional Release for income/release goals, Fan Raise + Release for a defined project-funding need when evidence supports testing fan participation, and Find a Professional when team gaps exist. Fan Raise must never be described as personal income and must never be treated as guaranteed funding. If the full project target is not supported, consider a smaller defined project-cost slice instead of excluding Fan Raise entirely.

Prepare a Bag Finder cross-market spread search that separates where to buy from where to resell. Use it to hunt distressed or wholesale-ish fixed-price sources and compare realistic exit channels using source-specific fees and completed-sale evidence. It does not purchase, contact sellers, or guarantee profit.

Prepare a fresh recheck brief for one owned Bag Finder deal candidate. Use this before re-verifying a WATCH, NEGOTIATE, or QUALIFIES candidate so price, availability, evidence, exit channels, and strategy status can be updated without losing history.

Prepare a category-specific Bag Finder search brief for lawful buy-low/sell-higher, resale, arbitrage, and flip opportunities such as real estate, vehicles, jewelry/watches, equipment, electronics, collectibles, inventory, websites/domains, and other lawful assets. This does not purchase, bid, contact sellers, or guarantee profit.

Prepare a strict item-level Bag Finder search for lawful distressed, motivated-seller, pawn, liquidation, closeout, wholesale-ish, or under-market inventory. This build rejects general sale/category pages and requires an exact item, exact asking price, direct current listing URL, and a realistic path to resale evidence before a lead can move beyond PASS. It does not purchase, contact sellers, or guarantee profit.

Prepare a verified Bag Finder search brief for finding outside opportunities that fit an owned profile. For broad or vague requests, do not call this until Bag Finder has first clarified what the user actually wants through guided conversational intake (goal, timing, amount/budget when relevant, starting assets, and key constraints). After the intent is clear, use this before web search and save only opportunities verified on current source pages.

Generate a time-limited SplitDaBag onboarding link for a public prospect. Call this only when the signed-in researcher explicitly asks to invite/onboard the prospect. The link lets the recipient claim the public research record after signing in, but it does not silently create, verify, or connect a Bag Finder profile.

Create or refresh a public-only Bag Finder research prospect for a person, artist, creator, business, brand, store, or professional who has not connected SplitDaBag yet. Use only public URLs and user-supplied context to prepare research queries. Do not treat user-supplied context as verified fact until current public sources confirm it.

Compare verified resale/exit channels for the same lawful asset. Bag Finder computes percentage and fixed selling costs, insured shipping/service costs, net proceeds, and evidence sufficiency so it can choose a realistic exit rather than simply the highest asking price.

Safely reconcile older active/watchlist deal records against the current saved strategy and current stored evidence. This maintenance action is downward-only: it may demote stale QUALIFIES/NEGOTIATE/WATCH records, dismiss recomputed PASS records, and backfill missing recheck schedules, but it never promotes a candidate. Use dry-run first when desired.

Record evidence for one of the founder-authored 10 artist-development lessons. This tool separates KNOWLEDGE completion from REAL-WORLD execution. A lesson is not ACHIEVED merely because the artist read it, clicked complete, or says they understand it. Use inspected analytics, campaign execution, DJ/promoter response, bookings, radio activity, merchandise results, business/safety preparation, and market expansion evidence as appropriate. If an earlier lesson needs repair, that repair outranks advancing the operating plan, although the artist may still read later lessons.

Record a verified recheck for one owned Bag Finder deal candidate, recompute evidence/economics under the current saved strategy, preserve price/status history, and promote or demote WATCH, NEGOTIATE, QUALIFIES, or PASS as the evidence changes. This never purchases or contacts a seller.

Run a deterministic Universal Deal Finder math and guardrail self-test. It validates exit-fee math, net proceeds, total project cost, profit, ROI, maximum buy, negotiation math, and the missing-economics no-promotion guard without saving or changing a deal candidate; standard MCP usage auditing still occurs.

Save and mathematically evaluate lawful resale/flip deal candidates after current public sources support the listing and material assumptions. When exit_channels are supplied, Bag Finder computes each exit's net proceeds after percentage/fixed fees, shipping/insurance, authentication/service and other exit costs, then underwrites the deal against a supported net exit rather than a retail asking price. Estimates are preliminary due diligence, not guaranteed profit or an appraisal.

Save outside opportunities that ChatGPT has already verified using current web sources. Only save opportunities supported by a real source URL. Saving the same source URL for the same profile updates the existing record instead of duplicating it, so this tool can overwrite saved opportunity state.