Pricing
Keep agent guidance in your hands. Choose manual skill management with Community, or automated learning with tailored deployment and governance with Enterprise.
Local skill management
For an individual managing skills locally.
Free
No time limit
No model key required
Manual processing runs locally without LLM usage.
No limits on local skills or rules.
Import, review, amend and publish agent skills yourself, with a record of where each rule came from.
Organisational controls
For organisations with specific control and deployment needs.
Bespoke
Priced following discussion
Bring your own model key
Software subscription; model usage billed separately by your provider.
Based on your deployment, assurance and support requirements.
Everything in Community, plus automated team learning and an agreed package of governance, integration and operational controls.
All editions keep your guidance portable. Enterprise adds automated learning, team distribution and controls agreed for your organisation.
Planned feature allocation. Enterprise scope and availability are agreed before purchase. On smaller screens, each capability is shown as a card.
| Capability | Community | Enterprise |
|---|---|---|
| Skill import, editing and version history | Included | Included |
| Rule lineage, retraction and restoration | Included | Included |
| Correction capture via MCP, HTTP and CLI | Included | Included |
| Correction review and target-skill selection | Manual | Automated, with review |
| Duplicate handling | Exact matches; human confirms | Semantic reconciliation |
| Portable SKILL.md, AGENTS.md and CLAUDE.md output | Included | Included |
| Publication and distribution | Local and manual Git | Scheduled Git, central management and agreed policy gates |
| MCP skill reads | Local | Scoped and centrally hosted |
| Capability | Community | Enterprise |
|---|---|---|
| Model API key | No key required | Bring your own key |
| LLM usage charges | No LLM usage | Billed separately by your provider |
| Automated distillation, scope judgement and skill selection | Not included | Included |
| Semantic conflict, supersession and cross-skill checks | Not included | Included |
| Authorial revision proposals and guarded application | Not included | Included |
| Continuous background compilation | Not included | Included |
| Contributors and access scoping | Single local operator | Shared contributors, teams, domains and enterprise RBAC |
| Personal mutes and overlays | Not included | Included |
| Graph exploration and reporting | Lightweight explorer | Full explorer, usage and activity |
| Capability | Community | Enterprise |
|---|---|---|
| Deterministic injection screening and regex PII sanitisation | Included | Included |
| Advanced injection screening and PII sanitisation | Not included | Included |
| Encrypted redaction custody and reveal controls | Not included | Included |
| Retention, erasure, audit export and webhooks | Not included | Agree requirements |
| Publication approvals and advanced policy gates | Not included | Agree requirements |
| SSO / OIDC | Not included | Included |
| SCIM and JIT user provisioning | Not included | By agreement |
| Deployment in your infrastructure | Local only | By agreement |
| Enterprise adapters, backup, recovery and support SLAs | Not included | By agreement |
You retain the published skills and supporting files. Installed skills remain usable without a live OMS connection; live MCP retrieval depends on OMS. Moving platforms may require integration changes and retesting.
Community processes corrections manually. To evaluate automated learning with your team’s agents and guidance, use “Discuss Enterprise” above. Hosted evaluation is planned as an explicit opt-in; Community processing stays local by default.