The context manager for enterprise agents
Get better outcomes from AI. Faster.
Reference Context is the shared workspace for the documents, rules, and guardrails your AI tools should follow - a dedicated MCP per task that’s fast to setup, versioned, auditable, and shared across your team.Keep your team in sync and your agents on task
PortableAny Agent or Model. Use the best agent, avoid vendor lockin, quickly move to new models as they launch.
PreciseCreate detailed context windows per project with auditable data sources and rules
CollaborativeShare context windows across teams to spread best practices and avoid stale or inaccurate data
FastPre-built rules customized to major Agents deliver state of the art performance out of the box. Do not waste time learning random prompt tricks.
Instantly get better AI outcomes
Bring every rule, source, and guardrail together in one workspace that powers the tools your team already uses. Set it up once and every service and agent your team already uses will just give better answers, faster, with fewer wasted tokens.You end up with
A project specific MCP server
Clear owners and review history
Instant tailored context delivery to any model
1
Create a Reference Context
A workspace that holds the data and rules for a project, wrapped in a project specific MCP server for use by any AI tool.2
Setup rules and guidance
Start from proven templates optimized for the or document the exact steps and boundaries your team already trusts.3
Attach the relevant data
Pull in documents, queries, specs, and data feeds from other MCPs - all with version tracking. Give AIs the right context upfront and spend less time debugging stale data.4
Get better answers
Connect the workspace to the chatbots, copilots, and workflows you already run with one click and see the improvement immediately, natively in the tools your team uses.How different teams stay aligned
Reference Context gives each team a single place to prepare the guidance their agents rely on. These are the everyday workflows customers ship in production.Customer Support Lead
Every reply links back to an approved source
Publish resolution guides once, reuse across Zendesk, Intercom, and Slack
Flag stale macros before they reach the queue
Track which teams acknowledged the latest policy change
Product Manager
Release notes that your agents can act on
Bundle specs, FAQs, and rollout dates into a single launch packet
Sync the packet to Jira, Notion, and internal chat
Archive prior versions so support can cite what changed
Compliance Director
Auditable guardrails without slowing the business
List approved vendors with the evidence legal requires
Require agents to acknowledge new guidance before acting
Export a full change log for regulators in minutes
Revenue Operations
Deals stay current across every CRM view
Update pricing narratives once and push everywhere reps sell
Highlight which assets were updated this week
Keep partner teams on the same set of numbers
IT & Security
Agents only use tools that passed review
Catalog approved integrations with owner and expiration dates
Share runbooks for incident response directly in chat
Alert teams when a certification or secret rotates
People Operations
Onboarding checklists that update themselves
Create role-based starter kits with the right docs and tutorials
Verify every manager confirmed the latest policy refresh
Give new hires a trusted place to ask and search
Real workflows Reference Context powers
Teams start with a trusted source of truth, then push that context into every agent, chat bot, and dashboard they run. Here’s what that looks like day to day.Product + Support
Launch a new product
1
Gather release brief, help center drafts, and pricing FAQs in a single workspace2
Tag each asset with the audience it applies to3
Ship updates to Zendesk, internal chat, and the product wiki in one publishRevenue Operations
Keep sales decks honest
1
Collect the latest metrics and customer stories from verified sheets2
Note who approved each claim and when3
Sync the package to Salesforce, Highspot, and agent promptsCompliance
Respond to an audit request
1
Filter policies by regulator and effective date2
Export the change log with reviewer comments3
Share a read-only workspace with the auditorPeople Operations
Onboard a new region
1
Clone the core onboarding kit and localize pay and policy details2
Assign tasks to regional leads with due dates3
Track completion and questions inside the workspaceFeatures
Focused Workspaces
Organize documents, tables, and MCP servers per workspace so every agent sees only the context that applies to the job at hand.Drag-and-drop folders from Drive, Confluence, or Notion with automatic source tagging.
Workspace-level document pinning keeps must-read briefs front and center.
Claude- or GPT-specific prompt packs (agents.md, claude.md, etc.) ship with every workspace.
Reference Tables & Live Queries
Blessed SQL snippets and LLM retrieval queries stay versioned and tested so agents never quote stale numbers.Run scheduled checks with drift alerts before data hits customer-facing flows.
Mark golden rows and expose them through a lightweight API for realtime use.
Document reviewer approvals and required plugins alongside each query.
Prompt Library & Plugin Picks
Browse a marketplace of prompt briefs tuned for leading agent platforms, then @-mention the right plugin or tool for your stack.Auto-detect the Claude Code plugins your team already uses and suggest vetted alternatives when you switch models.
Publish internal prompt templates with ownership, tags, and change history.
Link prompts to the workspaces and documents they depend on for easy reuse.
Rules & Tool Policies
Spell out exactly how agents should behave, which tools they must call, and which steps need human review.Compose style guides, escalation paths, and safety checklists with rich text blocks.
Require specific linters, test suites, or approval flows before an agent can mark work complete.
Attach tool rules to MCP servers so capabilities stay scoped to compliant projects.
Versioning & Reviews
Track every edit to docs, prompts, and rules with side-by-side diffs and reviewer sign-off.Compare prior guidance to the current state in one click for audit or rollback.
Request feedback from domain owners with automatic reminders when changes go stale.
Log which agents ran under which version to explain any output.
Collaboration & Tasks
Plan upcoming improvements and capture learnings so the next agent run benefits from the whole team.Assign owners and due dates to workspace tasks and let agents report when they’ve completed steps.
Collect experiment results, eval scores, and postmortems alongside the assets they reference.
Share workspace updates in Slack or email to keep cross-functional partners aligned.
Evaluations & Guardrails
Define the tests an agent must pass—accuracy checks, red-team prompts, or human QA—without writing code.Use point-and-click forms for non-technical reviewers to add new eval cases.
Gate production runs behind required test suites and capture results over time.
Flag regressions automatically and roll back to the last passing policy set.
Permissions & Audit Trail
Granular roles, access logs, and immutable history keep sensitive knowledge governed and reviewable.RBAC across workspaces, documents, tools, and evals with optional SSO enforcement.
Full activity log detailing who changed what and which agent consumed the update.
Export-ready reports for auditors or security reviews in seconds.
Pricing that scales with confidence
Start with a collaborative workspace and graduate to enterprise controls as your AI footprint expands.Starter
$79
per workspace / monthUnlimited playbooks
Real-time freshness
Audit trails
Dedicated success
Growth
Most Popular$249
per workspace / monthUnlimited playbooks
Real-time freshness
Audit trails
Dedicated success
Enterprise
Let’s talk
per monthUnlimited playbooks
Real-time freshness
Audit trails
Dedicated success