CT Accelerator MCP-Native · Knowledge Layer

ContextCT — makes every Cursor seat a Salesforce expert

A RAG-backed MCP server that grounds Composer, background agents, and BugBot in 10+ years of CT delivery IP — so Cursor ships correct Apex, LWC, and Flow from the first prompt.

The Documentation Gap

What no official docs cover

Salesforce docs teach the platform; Cursor docs teach the IDE. Neither teaches how to customize Salesforce with Cursor. ContextCT serves CT’s proven patterns exactly in that gap — on demand, inside the prompt.

Tribal Knowledge, Productized

10+ years of delivery IP, indexed

LSC data models, perfeCT mapping rules, governor-limit patterns, migration playbooks — curated into one governed vector index. Agents call resolve-project, then get-knowledge returns relevance-ranked context.

Why Cursor Wins

Evals are won on domain accuracy

Enterprise Salesforce evals fail on hallucinated Apex, not IDE features. Grounded agents produce code that compiles, deploys, and passes review — the difference between a stalled pilot and a team-wide rollout.

Any AI Client

Cursor · Claude · Agentforce

Standard MCP serves the IDE, cloud background agents, and CLI alike — published once through the Team MCP marketplace. No per-engineer setup, no knowledge drift.

Impact:  Cursor wins the enterprise
  • Higher eval win-rateGrounded agents pass Salesforce POCs that generic AI fails
  • Deeper seat adoptionDay-1 productive engineers — no ramp-up, no drift, no churn
  • Ecosystem moatPartner IP reachable only through MCP-native Cursor
CustomerTimes · Cursor-Powered LSC Delivery
CT Accelerator ContextCT · Zero-Friction Commerce

Knowledge that sells itself

The agent economy killed the sales call: skills sell like apps, marketplaces carry the paper, agents pay per call. ContextCT adopts all three — an adoption ladder with no decision meeting on it.

01 · Try — $0

Five minutes, no card, no call

One MCP install command in Cursor, 1,000 free retrievals a month. The first grounded answer is the pitch — the demo is the product, not a deck. Value lands before any invoice exists.

02 · Buy a Pack — $99/mo

Knowledge Packs, app-store style

À la carte expertise per team: LSC Data Models · perfeCT Mappings · Apex Limits · Flow Testing. Self-describing, cancel anytime — a card-swipe, not a committee. All-Access bundle: $150/SF admin seat.

03 · Buy Where You Already Buy

Ride paper that is already signed

Listed on AgentExchange — private offers, unified billing, auto-provisioning — and pushed via the Cursor Team MCP marketplace. Procurement inherits contracts the client already has: no new-vendor onboarding.

04 · Let the Agent Pay

x402-ready: agents buy their own context

Per-call micropayments under a pre-authorized cap (x402 / MPP sessions). No subscription, no account — a wallet. The endgame: knowledge procured autonomously by the agent that needs it.

No decision meeting:  every step self-serve & reversible
  • Value before invoiceThe first answer lands before the first dollar
  • Packs, not platforms$99 is a card-swipe — a unit buyers already understand
  • Paper already signedMarketplace billing + agent wallets — zero new procurement
Patterns: AgentExchange unified billing & private offers · SKILL.md skill marketplaces · x402 / MPP agentic payments — 2026.
CustomerTimes · Cursor-Powered LSC Delivery
CT Accelerator ContextCT · Digital Labor Pricing

Price the work, not the seat

Seat comparisons undersell it: ContextCT doesn’t displace license spend — it displaces implementation labor. Sell it as digital labor, billed in the currency Salesforce buyers already budget.

The Real Baseline — implementation economics
ContextCT seat / month$150
One SF consultant hour$150–250
One consultant day$1.6–2k
Senior LSC consultant / month$25–40k
Off the chart: a typical OCE/Veeva → LSC migration services line runs $500k–$2M. Against that budget, the knowledge layer is a rounding error — with leverage on every hour of it.
The Model — Digital Labor Credits
  • One prepaid currency — credit packs with Flex-Credits mechanics: the procurement motion SF buyers already have approvals for
  • Spend on agent work — retrievals and token-metered context, as agents build
  • Spend on outcomes — deliverables bill only on acceptance: AI-assisted PR merged, object mapping passed UAT
  • Convert to humans — the same credits redeem CT expert hours when agents hit their limits: one fungible workforce
The unlock: bills to the services budget — the six-figure line the client already planned — not to a new software line fighting for approval.
FDE packages:  skilled Cursor engineers, embedded
  • 0.5 FTE — from $15k/moFractional FDE: reviews, unblocking, knowledge curation
  • 2 FTE — from $55k/moEmbedded squad: builds with your team, runs the agent fleet
  • 5 FTE — from $125k/moFull program: end-to-end delivery, credits bundled
Why it closes:  risk moves to CT, budget stays put
  • Zero-risk entryOutcomes bill on acceptance — shipped work, not lookups
  • No new budget lineConsumes the implementation budget already approved
  • Built-in retentionCredits & FDE hours roll into the next phase
Consultant rates: typical senior Salesforce/LSC market ranges, 2026. Credit denominations to be validated with pilot accounts.
CustomerTimes · Cursor-Powered LSC Delivery