mcp-scraper 0.21.0 → 0.21.2

package/docs/specs/unified-credit-and-scheduled-execution-billing-spec.md

@@ -1,7 +1,7 @@
 # Unified Credits, Integrations, and Scheduled Execution Billing
 
-- Status: Proposed canonical implementation spec
-- Spec date: 2026-07-13
+- Status: Released v1 contract plus explicitly deferred hardening roadmap
+- Spec date: 2026-07-14
 - Primary repos: mcp-scraper, mcp-scraper-scheduler, mcpscraper-web, mcpscraper-sdk
 - Commercial decision: Scheduled Actions is a product capability, not a separate subscription
 - Accounting decision: MCP Scraper's main credit ledger is the only debit authority
@@ -10,21 +10,29 @@
 
 MCP Scraper will use one visible usage currency: Credits.
 
-Integrations, bulk connected-data movement, scheduled execution, model inference, and memory ingestion will all consume the same wallet. Scheduled Actions will no longer be sold as a separate $10-per-month product with a 1,000-run quota.
+Scheduled Actions uses the same Credit wallet as the rest of MCP Scraper. It is no longer sold as a separate $10-per-month product with a 1,000-run quota. Integrations themselves are included with paid plans: connecting, reconnecting, direct reads, enabled actions, supported exports, and connection-to-Memory snapshots do not currently create a separate integration-operation debit.
 
 The replacement contract is:
 
 1. Every paid MCP Scraper plan includes access to integrations and scheduling.
-2. A customer without a recurring plan may activate metered features with a minimum $10 credit purchase once pay-as-you-go activation exists.
-3. Each scheduled occurrence that successfully reserves Credits and begins execution costs a small fixed execution charge plus the exact work performed.
-4. OpenRouter-backed AI is settled from OpenRouter's actual reported cost, multiplied by 1.5, then converted to Credits.
-5. Token prices may be shown as current per-million-token estimates in Credits, but the reported OpenRouter cost is the invoice truth.
-6. Active connections have a monthly Credit minimum that is offset by their actual eligible connection usage. This pays for OAuth, refresh, provider infrastructure, and idle connection overhead without charging a confusing separate integration subscription.
-7. Every execution path—the public MCP, REST API, SDK, Mastra, background sync, and frontend—uses the same billing adapter and produces the same itemized receipt.
+2. OAuth connect and reconnect cost 0 Credits and there is no per-connection subscription or active-connection minimum.
+3. Each scheduled occurrence that passes the balance check and begins execution costs 75 Credits exactly once.
+4. A deterministic `connection_sync` occurrence has no model charge beyond that 75-Credit base.
+5. An agent-mode occurrence additionally settles OpenRouter's actual reported model cost, multiplied by 1.5, then converted to Credits.
+6. Token prices may be shown as current per-million-token estimates in Credits, but OpenRouter's reported cost is the settlement truth.
+7. Separately metered scraper, browser, transcription, video-analysis, and similar tools keep their published rate-card prices when a schedule calls them.
+8. The proposed 5/10/15/25-Credit connected-operation tariff, per-record/byte tariff, and 20,000-Credit active-connection minimum are deferred and are not enabled.
 
-This design encourages usage: paid users can immediately connect and schedule work, higher-usage customers naturally consume more Credits, and there is no artificial 1,000-run cliff or add-on product to understand.
+This design encourages usage: paid users can immediately connect and use accounts without per-operation friction, while recurring automation and model work draw from the shared Credit balance. There is no artificial 1,000-run cliff or add-on product to understand.
 
-## What exists today
+### Released v1 implementation boundary
+
+- Authorization currently performs an idempotent wallet balance check, not a hard balance reservation.
+- The server-side default maximum is 5,000 Credits per run and is environment-configurable; it is not yet editable in the customer UI.
+- A missing authoritative model cost moves the schedule to `cost_pending` and pauses it. Automated recovery/reconciliation is still a hardening item.
+- Connected-operation receipts, user-editable monthly caps, auto-refill, and active-connection minimums remain future work unless a later commercial decision explicitly enables them.
+
+## What was found before the v1 cutover
 
 The originally intended scheduler economics are recognizable in the code, but the implementation is not safe enough to keep.
 
@@ -38,9 +46,9 @@ The originally intended scheduler economics are recognizable in the code, but th
 
 So the answer to “was $10 meant to open the feature and then AI cost was marked up 50%?” is: substantially yes. The entitlement and the 50% markup both exist.
 
-### Current billing defect
+### Pre-cutover billing defect
 
-The effective path does not reliably implement that intended price:
+The pre-cutover path did not reliably implement that intended price:
 
 1. schedule-execute invokes both meterLlmUsage and billPassThroughCost for the generated result.
 2. Both paths can call recordCost.
@@ -50,7 +58,7 @@ The effective path does not reliably implement that intended price:
 6. The current metering client is fail-open, so missing configuration or a failed internal request can instead produce no debit.
 7. The current call uses top-level usage and provider metadata. For a multi-step agent result, this describes the last step, not necessarily the complete run. Earlier model steps can be omitted.
 
-The live scheduler environment has the internal MCP Scraper URL, internal debit secret, and OpenRouter key configured. This means the path is wired in production configuration; it is not merely dead local code.
+The scheduler environment had the internal MCP Scraper URL, internal debit secret, and OpenRouter key configured, so this was a real production risk rather than dead local code. The v1 cutover removed the duplicate/fail-open path and now aggregates model-step costs before one settlement.
 
 ### Required immediate interpretation
 
@@ -58,25 +66,25 @@ Do not assume existing scheduled-agent charges equal a clean 1.5 markup. Before
 
 ## Canonical Build Spec
 
-- **Request:** Replace separate Scheduled Actions billing with a unified Credit wallet; correctly settle connected-service operations, deterministic syncs, and all-step OpenRouter costs; expose clear per-run budgets and receipts on every product surface.
+- **Request:** Replace separate Scheduled Actions billing with the shared Credit wallet; charge one 75-Credit occurrence base and correctly settle all-step OpenRouter cost at 1.5; keep direct integration use included.
 - **Lane:** Mastra Agentic Development plus MCP/API product billing.
 - **Capability:** Unified Metered Execution.
 - **Audience:** MCP Scraper customers, scheduled-agent users, MCP clients, SDK users, operators, finance, and support.
-- **User-Facing Outcome:** A user connects a service, optionally moves its data into memory, creates a schedule, sees an estimate and spending limit, and pays one itemized Credit charge for each completed or partially completed unit of work.
+- **User-Facing Outcome:** A user connects a service, uses it directly without a separate operation fee, optionally moves a bounded result into Memory, and pays 75 Credits plus applicable model/tool usage when recurring work executes.
 - **Target Runtime:** Main MCP Scraper API and credit ledger; Vercel-hosted Mastra scheduler; Inngest workers; Nango connections and Proxy/functions; private Blob artifacts; public MCP; npm/SDK clients.
 - **Inputs:** Authenticated identity, connection ID, tool or sync request, scheduled occurrence, model selection, per-run and monthly Credit limits, actual provider results, OpenRouter usage metadata, bytes and records delivered.
 - **Agents:** Existing Scheduled Action Agent for interpretive work. Deterministic connection_sync workers remain non-agentic. Both use the same billing context.
 - **Workflows:** Connect; direct read/write; bulk export; memory import; schedule estimate; occurrence claim; Credit authorization; execution; incremental settlement; final settlement; receipt; retry/reconciliation.
 - **Tools:** Existing generic connected-service tools, provider-native tools, schedule tools, bulk export/import tools, credits_info, a new schedule estimator, and a shared internal billing adapter.
 - **MCP:** Credit behavior is transport-independent. Tool descriptions state charge categories and estimates, results include receipt summaries, and exact provider tools remain protected by tenant ownership, action switches, and grants.
-- **Signals:** Billing authorization created; low balance; budget threshold; cost pending; settlement complete; insufficient Credits; automatic refill complete/failed; connection minimum due; connection billing paused.
+- **Signals:** Billing authorization created; low balance; budget threshold; cost pending; settlement complete; and insufficient Credits. Auto-refill and connection-minimum signals belong only to the deferred roadmap.
 - **Background Tasks:** Usage aggregation, connection-minimum settlement, pending-cost reconciliation, low-balance notifications, automatic refill, legacy subscription migration, receipt retention, and rate-snapshot refresh.
 - **Guardrails:** Integer microcredit arithmetic, idempotent charges, one debit authority, no model-visible billing credentials, fail closed before externally costly work, action approval independent of price, no prompt or provider credential in billing events.
 - **Observability:** Itemized receipt per run; raw vendor cost; multiplier; model/provider; tokens; bytes; changed records; connection and schedule IDs; idempotency keys; authorization, settlement, and refund state.
 - **Enforcement:** A call cannot bypass the billing adapter by using Mastra, the generic MCP bridge, direct REST, SDK, Nango, or a background job. Duplicate occurrence or HTTP retry cannot debit twice.
 - **Persistence:** The main MCP Scraper/Turso credit ledger owns balances and debits. Scheduler Postgres owns schedule state, occurrence state, and a mirrored receipt reference, not the authoritative wallet.
 - **Deployment:** Main API/ledger first, then scheduler and connection adapters, then public MCP, frontend, SDK/npm/MCPB, generated docs, Stripe migration, and Vercel production.
-- **Done Contract:** A production scheduled run authorizes a budget, records all model steps and connected-service work once, settles one exact Credit receipt, releases unused Credits, and presents the same receipt through the UI, MCP, API, and SDK.
+- **Done Contract:** A production scheduled run performs one balance authorization, charges its 75-Credit base once, aggregates every OpenRouter model step once, and exposes the same pricing contract through the UI, MCP, API, and SDK. Hard reservations and full receipt UX remain the next hardening phase.
 - **Receipt Requirements:** Code and deployment SHAs; rate-table revision; Stripe migration result; test output; production Vercel IDs; npm/MCPB versions; sample agent and deterministic-sync billing receipts; reconciliation report; public pricing and tool-description parity.
 
 ## Product contract in plain language
@@ -85,7 +93,7 @@ Do not assume existing scheduled-agent charges equal a clean 1.5 markup. Before
 
 Scheduled Actions remains a first-class page and capability. It is not a separate product, subscription, or quota bundle.
 
-Users with an active paid plan can create schedules immediately. Pay-as-you-go accounts can create schedules after a minimum $10 wallet activation is released. The $10 is a Credit purchase, not an expiring scheduler access fee.
+Users with an active paid plan can create schedules immediately. A future pay-as-you-go activation may use a minimum $10 Credit purchase, but that path is not part of the released v1 contract. If released, the $10 must be a wallet purchase rather than an expiring scheduler access fee.
 
 Existing scheduled-action subscriptions are migrated; customers do not keep paying a separate scheduler subscription.
 
@@ -93,11 +101,12 @@ Existing scheduled-action subscriptions are migrated; customers do not keep payi
 
 A scheduled occurrence pays:
 
-    fixed execution charge
-    + connected-service work
-    + bulk records and bytes delivered
-    + model usage
-    + memory processing
+    75-Credit execution charge
+    + 1.5 × actual OpenRouter cost when agent mode uses a model
+    + any separately metered scraper, browser, transcription,
+      video-analysis, or other published rate-card tools it calls
+
+Direct connected-service reads, enabled actions, supported exports, and bounded Memory snapshots do not add a separate integration-operation line item today.
 
 Internal pagination, retries, leases, and orchestration steps are not separate “runs.” A retry of the same occurrence cannot pay the fixed charge again.
 
@@ -162,36 +171,34 @@ Auto-refill is off by default. The customer chooses a threshold and a monthly ma
 
 This is intentionally less generous than subscription grants. Plans remain the best value; pay-as-you-go exists for activation and overflow.
 
-## Canonical rate table
+## Released v1 rate table
 
 All rates live in one versioned server-side registry. Frontends, MCP descriptions, SDKs, and docs consume generated projections from that registry.
 
 ### Connection operations
 
-Initial launch rates:
+Released v1 rates:
 
 | Billing class | Charge |
 |---|---:|
 | OAuth connect, reconnect, or consent callback | Free |
-| Successful bounded read | 5 Credits |
-| Successful low-complexity write | 10 Credits |
-| Successful medium-complexity write | 15 Credits |
-| Successful high-risk or high-complexity approved write | 25 Credits |
-| Start a bulk export or sync slice | 25 Credits |
-| Delivered or changed record | 2 Credits each |
-| Transferred content | 10 Credits per started MiB |
-| Provider or MCP Scraper validation failure before provider work | Free |
-| Duplicate page, unchanged delta record, or idempotent replay | Free |
+| Direct approved read | Included |
+| Direct enabled action | Included |
+| Supported connected-service export | Included |
+| Bounded connection-to-Memory snapshot | Included |
+| Active connection minimum | None |
 
-Tool manifests classify each action; a caller cannot choose a cheaper class. The provider-policy review owns read/write/risk classification.
+Action safety remains independent of price: ownership, live inventory, source-controlled policy, the account action switch, and exact scheduled grants still apply.
 
-Charge for the outcome, not raw upstream requests. For example, a 500-item export that requires ten provider pages is one bulk base plus 500 delivered records, not ten separate read charges.
+### Deferred connected-operation pricing options (not enabled)
 
-A bounded read is limited to one logical object or at most 100 summary records and 1 MiB of structured output. Requests above either threshold use the bulk contract. A provider tool cannot evade bulk rates merely by accepting a very large page size.
+Earlier design work proposed 5/10/15/25-Credit read/write classes, a 25-Credit bulk base, per-record charges, and per-MiB charges. Those values are retained only as research for a later commercial decision. They are not present in the production registry and must not appear as current pricing in UI, MCP descriptions, SDKs, or support answers.
 
-### Active connection minimum
+### Deferred active connection minimum (not enabled)
 
-Each active connection has a 20,000-Credit minimum per 30-day billing period, prorated to the time it is active.
+An earlier proposal described a 20,000-Credit minimum per active connection per 30-day billing period. The released product has no active-connection minimum, no first-connection waiver clock, and no connection billing-pause flow. Enabling any such minimum requires a new explicit commercial decision, customer notice, and a separate release.
+
+The remainder of this subsection records the deferred proposal for future evaluation only:
 
 One tenant-owned credential/account row is one connection. Two Google accounts, two Slack workspaces, or two WordPress sites are separate connections. Disconnected and billing-paused credentials do not accrue a new minimum; active and reauthorization-grace credentials accrue only for the active seconds in the period.
 
@@ -249,20 +256,14 @@ It also allows a light schedule to remain inexpensive. A schedule running at 8 A
 An agent occurrence settles:
 
     75-Credit occurrence base
-    + provider tool charges
-    + connected data records/bytes
     + OpenRouter actual cost times 1.5
-    + memory ingestion or artifact processing
+    + any separately metered published tools invoked by the run
 
 #### Deterministic connection sync
 
 A connection_sync occurrence settles:
 
     75-Credit occurrence base
-    + 25-Credit bulk slice base when a slice starts
-    + changed records
-    + transferred bytes
-    + memory embedding/ingestion processing
 
 There is no model charge unless that deterministic pipeline actually invokes a model.
 
@@ -298,24 +299,22 @@ Assume one MiniMax M3 run uses 100,000 input tokens, 20,000 output tokens, and o
 | Line item | Credits |
 |---|---:|
 | Scheduled occurrence | 75 |
-| Connected-service read | 5 |
+| Connected-service read | Included |
 | Estimated model input | 300 |
 | Estimated model output | 240 |
-| Estimated total | 620 |
+| Estimated total | 615 |
 
-The model lines settle from the actual reported OpenRouter cost, so 620 is an estimate until the receipt closes.
+The model lines settle from the actual reported OpenRouter cost, so 615 is an estimate until the receipt closes.
 
 #### Deterministic bulk sync
 
-Assume a sync delivers 500 changed records and 50 MiB:
+Assume a deterministic connection sync delivers 500 changed records and 50 MiB without calling a separately metered tool:
 
 | Line item | Credits |
 |---|---:|
 | Scheduled occurrence | 75 |
-| Bulk slice | 25 |
-| 500 changed records | 1,000 |
-| 50 MiB transferred | 500 |
-| Total before memory processing | 1,600 |
+| Connected records and transfer | Included |
+| Total | 75 |
 
 ### Existing non-scheduler AI classes
 
@@ -347,24 +346,16 @@ The UI may place raw dollar rates in a secondary tooltip or expandable explanati
 
 ## Per-run and monthly budgets
 
-Every schedule has:
-
-- maxCreditsPerRun, default 5,000 Credits;
-- maxCreditsPerMonth, default 50,000 Credits;
-- lowBalanceBehavior, initially pause;
-- optional model allowlist;
-- optional maximum model steps and output tokens;
-- an estimate produced before activation.
-
-The user can edit these limits without recreating the schedule.
+Released v1 has a server-side `maxCreditsPerRun` default of 5,000 Credits and pauses on insufficient balance or unresolved model cost. The cap is environment-configurable and is not yet user-editable. User-defined monthly caps, model allowlists, auto-refill behavior, and a pre-activation estimator remain future work.
 
 Before each occurrence:
 
-1. Calculate a conservative reservation from the per-run cap and available balance.
-2. If the monthly cap has insufficient headroom, skip with no run-base charge.
-3. If the wallet cannot reserve the required amount, skip with no run-base charge and notify the owner.
-4. Stop further model steps and provider mutations before exceeding the reserved cap.
-5. Release unused reserved Credits after settlement.
+1. Perform an idempotent balance authorization before model execution.
+2. If the wallet cannot cover the maximum, skip without a run-base charge.
+3. Settle the 75-Credit base once plus all authoritative model-step cost.
+4. If authoritative model cost is missing, move to `cost_pending` and pause later occurrences.
+
+Hard reservations, incremental budget stopping, monthly caps, and automatic release of unused reserved balance remain hardening work.
 
 The current very large max-output-token ceiling cannot function as the spending guard. Budget enforcement must occur during execution, including after each model step.
 
@@ -374,28 +365,26 @@ The current very large max-output-token ceiling cannot function as the spending
 
     Inngest or scheduler trigger
       -> idempotently claim identity + schedule + scheduledFor
-      -> ask main billing service for a Credit reservation
-      -> receive opaque billing context and reservation
+      -> ask main billing service for an idempotent balance authorization
+      -> receive an opaque billing context
       -> execute exact schedule grants
-      -> attach billing context to internal child calls
-      -> accumulate provider, record, byte, model, and memory line items
-      -> settle actual integer mc once
-      -> release unused reservation
-      -> persist itemized receipt reference
-      -> expose the same receipt on API, MCP, SDK, and UI
+      -> aggregate every OpenRouter model-step cost
+      -> settle the 75-Credit base plus model amount once
+      -> persist billing event references
 
 The billing context travels in an internal authenticated transport header or server-side execution context. It is never put in the model prompt, tool arguments, provider payload, or public result.
 
+Hard reservations, incremental child-call line items, and full customer-facing itemized receipts are the next hardening phase rather than released-v1 claims.
+
 ### Direct non-scheduled flow
 
-Direct MCP, REST, or SDK work follows the same abbreviated path:
+Direct connected-service MCP, REST, or SDK work is included and still follows the same tenant ownership and permission checks. If that work invokes a separately metered public tool, the existing tool-specific billing path applies:
 
-    authorize maximum charge
-      -> execute tenant-owned exact tool
-      -> settle outcome
-      -> return result plus receipt summary
+    validate tenant-owned exact tool
+      -> execute the connected-service operation without an integration debit
+      -> separately meter only published scraper/media/browser classes used
 
-The generic MCP bridge and a provider-native MCP tool must produce the same billing class for the same operation.
+The generic MCP bridge and a provider-native MCP tool must not silently produce different integration pricing for the same operation.
 
 ### Multi-step model accounting
 
@@ -430,9 +419,9 @@ The main MCP Scraper billing service is the only component allowed to mutate the
 
 The existing internal memory AI-debit endpoint becomes a compatibility adapter during migration, then routes through the unified billing service. It may not apply an extra 3.0 multiplier to a scheduled_agent_llm event.
 
-## Billing data model
+## Billing data model and hardening roadmap
 
-Names are illustrative; final migrations follow existing repository naming.
+The released v1 authorization and billing-event tables are smaller than the target below. Names are illustrative; the following reservation, receipt, and connection-period structures describe future hardening and must not be treated as already released.
 
 ### billing_events
 
@@ -515,7 +504,7 @@ The canonical occurrence key is:
 
     schedule:{identityId}:{scheduleId}:{scheduledForIso}
 
-It owns at most one occurrence-base event. The base event is created only after the claim obtains a Credit reservation and execution begins. Worker retries, leases, and resumed slices reference that event.
+It owns at most one occurrence-base event. In released v1 the base event is created only after the claim obtains its balance authorization and execution begins. Future hard-reservation work will retain the same key. Worker retries, leases, and resumed slices reference that event.
 
 ### Line-item keys
 
@@ -534,7 +523,7 @@ The main ledger rejects a second settlement for the same key and returns the exi
 
 - Work completed before a provider or user error remains chargeable.
 - No provider-work line is charged for validation failure before the provider call.
-- Changed records already delivered remain charged if a later page fails.
+- Included connected-service records and bytes do not create their own charge in released v1.
 - Unused reservation is always released.
 - The 75-Credit platform base is refunded when MCP Scraper itself fails before useful execution.
 - The base is not refunded for invalid customer configuration, expired OAuth after a valid start, provider outage after dispatch, or user-authored agent failure unless support policy grants a courtesy refund.
@@ -597,6 +586,8 @@ Refunds append compensating ledger events. Settled rows are never silently edite
 
 ## Public API, MCP, and SDK contract
 
+Released v1 publishes the 75-Credit occurrence base and 1.5-times actual OpenRouter rule across the hosted MCP, npm packages, generated SDK contracts, marketing pages, and scheduler status. The richer fields below are the follow-on receipt and budget roadmap.
+
 ### Existing tools to extend
 
 credits_info adds:
@@ -706,10 +697,8 @@ The deterministic worker:
 
 - claims the same occurrence key;
 - authorizes its budget;
-- reports one bulk-slice base;
-- reports only changed/delivered records;
-- reports content bytes once per stored artifact;
-- reports memory processing through existing explicit classes;
+- charges the 75-Credit occurrence base once;
+- does not add record, byte, or connected-operation charges in released v1;
 - checkpoints before settlement so retry behavior is deterministic.
 
 The worker never invokes a model merely to paginate or transfer content.
@@ -724,14 +713,7 @@ Under every paid plan:
 
 > Integrations and scheduled work included. Usage draws from your shared Credit balance.
 
-Add a concise rate drawer containing:
-
-- active connection minimum;
-- read/write/bulk rates;
-- 75-Credit scheduled occurrence;
-- current model estimates in Credits;
-- “actual OpenRouter cost plus 50%” disclosure;
-- expiry, refill, and budget behavior.
+Show the 75-Credit scheduled occurrence, the “actual OpenRouter cost plus 50%” disclosure, and the fact that direct integration use is included. Do not show the deferred active-connection minimum or per-operation tariff as current pricing.
 
 ### Integrations page
 
@@ -740,10 +722,6 @@ Each card shows:
 - connected account count;
 - real callable read/action inventory;
 - action switch;
-- current billing-period usage;
-- minimum progress;
-- next settlement;
-- billing state;
 - a link to import/schedule.
 
 OAuth authorization itself remains free.
@@ -782,6 +760,10 @@ Users can filter by connection, schedule, model, provider, and date.
 
 ## Scheduled Actions subscription migration
 
+### Live migration result
+
+The legacy live-mode Stripe price `price_1ToXcHS8aAcsk3TGf3cW7zHx` still exists as an active $10 monthly price, but a read-only all-status audit on 2026-07-14 found zero subscriptions. The old checkout route now returns `410 Gone`, so no customer cancellation, proration, or compensating migration Credit is currently required. The price may be archived separately as Stripe catalog cleanup; archiving it is not required for runtime cutover and was not performed automatically.
+
 ### Customer treatment
 
 1. Stop new scheduler-subscription checkout after the unified billing path is proven.
@@ -854,11 +836,11 @@ Issue compensating Credits for defensible overcharges. Do not retroactively debi
 - Enable per-run/month caps and cost-pending reconciliation.
 - Expand to production cohorts after receipt review.
 
-### Gate 5 — deterministic sync and direct connection billing
+### Gate 5 — deterministic sync and direct connection billing (partially deferred)
 
-- Route connection_sync and bulk export through the same billing context.
-- Then route generic MCP, REST, and SDK calls.
-- Start active-connection minimums only after a full shadow period and customer notice.
+- Route `connection_sync` through the shared 75-Credit occurrence context.
+- Keep generic MCP, REST, SDK, exports, and snapshots included in released v1.
+- Reconsider connected-operation tariffs or active-connection minimums only after a new explicit commercial decision, shadow period, and customer notice.
 
 ### Gate 6 — commercial migration
 
@@ -908,13 +890,10 @@ Issue compensating Credits for defensible overcharges. Do not retroactively debi
 ### Connection tests
 
 - OAuth callback is free.
-- Bounded reads and approved writes receive correct server-owned classes.
+- Direct reads, enabled actions, exports, and bounded Memory snapshots do not create an integration-operation debit.
 - Internal pagination is not separately billed.
-- Unchanged incremental records are not charged.
-- Bytes are charged once per artifact.
-- Direct MCP, REST, SDK, Mastra, and background calls produce parity.
-- Connection minimum offsets eligible usage and prorates correctly.
-- First-connection waiver and billing-pause grace work.
+- Direct MCP, REST, SDK, Mastra, and background calls publish the same included-use rule.
+- Deferred operation tariffs and active-connection minimums are absent from the production registry and public copy.
 
 ### Ledger tests
 
@@ -961,7 +940,7 @@ Receipts must be sanitized. Do not include prompts, connection tokens, API keys,
 - Separate scheduling entitlement or 1,000-run quota still enforced on any surface.
 - Frontend, hosted MCP, SDK/npm, MCPB, or public docs showing different rates.
 - A fixed token rate described as authoritative rather than an estimate.
-- Connection minimum enabled without customer notice, waiver, and billing-pause behavior.
+- Any connected-operation tariff or active-connection minimum exposed without a new approved commercial release.
 - Commercial migration without a reconciliation and compensating-credit policy.
 
 ## Decisions locked by this spec
@@ -972,10 +951,10 @@ Receipts must be sanitized. Do not include prompts, connection tokens, API keys,
 - scheduled_agent_llm uses a 1.5 multiplier.
 - Model token tables are estimates.
 - The fixed scheduled-occurrence charge is 75 Credits.
-- The launch active-connection minimum is 20,000 Credits per 30 days, offset by eligible connection usage.
+- Direct integration use is included; the proposed 5/10/15/25 tariffs and 20,000-Credit active-connection minimum are not enabled.
 - OAuth is free.
 - Main MCP Scraper owns the only mutable Credit ledger.
-- Every product and execution surface produces a shared receipt.
+- Every product and execution surface publishes the same price contract; shared itemized receipt UX remains hardening work.
 - Action authorization remains separate from billing.
 
 ## Decisions that may change only through a rate revision

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "mcp-scraper",
-  "version": "0.21.0",
+  "version": "0.21.2",
   "description": "MCP server for MCP Scraper web intelligence tools",
   "type": "module",
   "main": "./dist/index.cjs",
@@ -76,7 +76,7 @@
     "hono": "^4.12.18",
     "inngest": "^4.4.0",
     "jose": "^6.2.3",
-    "mcpscraper-memory-tools": "^1.5.0",
+    "mcpscraper-memory-tools": "^1.5.1",
     "p-limit": "^7.3.0",
     "papaparse": "^5.4.0",
     "playwright": "^1.44.0",

package/dist/chunk-L4LMSXLO.js

@@ -1,7 +0,0 @@
-// src/version.ts
-var PACKAGE_VERSION = "0.21.0";
-
-export {
-  PACKAGE_VERSION
-};
-//# sourceMappingURL=chunk-L4LMSXLO.js.map

package/dist/chunk-L4LMSXLO.js.map

@@ -1 +0,0 @@
-{"version":3,"sources":["../src/version.ts"],"sourcesContent":["export const PACKAGE_VERSION = '0.21.0'\n"],"mappings":";AAAO,IAAM,kBAAkB;","names":[]}