DOC-1867 AI Gateway #477
Open
DOC-1867 AI Gateway #477
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Contributor
|
Important Review skippedAuto incremental reviews are disabled on this repository. Please check the settings in the CodeRabbit UI or the You can disable this status message by setting the
📝 WalkthroughWalkthroughAdds extensive AI Gateway documentation and navigation: new AI Gateway index, overview, quickstart, architecture, CEL routing cookbook, observability (logs/metrics), migration guide, MCP aggregation/orchestration guide, administrator and builder setup guides, discovery/connect guides, many IDE/tool integration pages (Claude Code, Cline, Continue.dev, Cursor, GitHub Copilot), MCP Server Patterns/nav updates, a BYOC support partial, and renames the AI Agents index to "Agentic AI". All changes are documentation-only. Sequence Diagram(s)sequenceDiagram
participant Client
participant AI_Gateway as "AI Gateway"
participant MCP_Server as "MCP Server"
participant LLM_Provider as "LLM Provider"
participant Observability as "Observability"
Client->>AI_Gateway: Send request (headers, rp-aigw-id, body)
AI_Gateway->>Observability: Log request + metrics
AI_Gateway->>AI_Gateway: Apply policies (CEL, rate/spend limits, routing)
AI_Gateway->>MCP_Server: Discover/Invoke tools (deferred load / orchestrator)
MCP_Server-->>AI_Gateway: Tool result(s)
AI_Gateway->>LLM_Provider: Forward transformed request (provider pool / fallback)
LLM_Provider-->>AI_Gateway: Response (tokens, cost)
AI_Gateway->>Observability: Log response, costs, routing decisions
AI_Gateway-->>Client: Consolidated response (LLM + MCP tool outputs)
Estimated code review effort🎯 3 (Moderate) | ⏱️ ~25 minutes Possibly related PRs
Suggested reviewers
🚥 Pre-merge checks | ✅ 5✅ Passed checks (5 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. ✨ Finishing touches🧪 Generate unit tests (beta)
Comment |
✅ Deploy Preview for rp-cloud ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
c4milo
reviewed
Jan 6, 2026
c4milo
reviewed
Jan 6, 2026
|
|
||
| This can reduce token usage significantly (for example, 80-90% depending on how many servers/tools are configured). | ||
|
|
||
| == Observability |
Member
There was a problem hiding this comment.
I haven't gotten to the observability implementation yet, probably next week.
|
|
||
| This is central to governance: You can see and control usage by gateway boundary (for example, by team, environment, customer, or product). | ||
|
|
||
| *REVIEWERS: Where do those metrics appear in the UI, or how does a user validate observability after setup?* |
micheleRP
force-pushed
the
DOC-1867-Document-feature-AI-Gateway-help-cloud-team-polish-clean-up
branch
from
9aef822 to
b28bea7
Compare
c4milo
reviewed
Jan 6, 2026
| ---- | ||
| { | ||
| "claude.mcpServers": { | ||
| "redpanda-ai-gateway": { |
Member
There was a problem hiding this comment.
This configuration should use http as transport. For example, to configure our internal MCP gateway in Claude Code, it would look roughly like this:
claude mcp add --transport http redpanda-aigateway https://gw.ai.panda.dev/mcp --header "Authorization: Bearer YOUR_API_KEY" --header "rp-aigw-id: GATEWAY_ID"
Member
There was a problem hiding this comment.
There is also a way to disallow people from using other MCP servers. I'm going to have to dedicate some time to documenting and testing all the integrations with popular AI coding agents.
# Conflicts: # modules/ROOT/nav.adoc
Added 7 new documentation files for AI Gateway: - what-is-ai-gateway.adoc: Overview, problem/solution framing, common patterns - quickstart-enhanced.adoc: Step-by-step quickstart with time markers - observability-logs.adoc: Request logs, filtering, and debugging - observability-metrics.adoc: Dashboards, analytics, and cost tracking - migration-guide.adoc: Safe migration from direct provider integration - cel-routing-cookbook.adoc: CEL routing patterns with examples - mcp-aggregation-guide.adoc: MCP aggregation and orchestration All files follow Redpanda documentation standards: - Sentence case headings - Imperative verbs for action headings - AsciiDoc format - Comprehensive placeholders for product-specific details Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Add personas, learning objectives, and prerequisites to all AI Gateway documentation pages. Remove DRAFT prefixes from titles and time estimates from quickstart. Fix passive voice in multiple locations. Changes: - Add page-personas attributes to all 7 files - Add learning objectives in ABCD format - Add prerequisites sections where missing - Remove "DRAFT:" from all page titles - Remove time estimates from quickstart-enhanced.adoc - Fix passive voice constructions - Improve page descriptions - Preserve all placeholder comments for future content Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
micheleRP
force-pushed
the
DOC-1867-Document-feature-AI-Gateway-help-cloud-team-polish-clean-up
branch
from
1297ef0 to
49a573f
Compare
Provide admin and user guides for configuring Claude Code, Cline, Continue.dev, Cursor IDE, and GitHub Copilot to work with AI Gateway, enabling centralized LLM routing and MCP tool aggregation. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
…oud-team-polish-clean-up # Conflicts: # modules/ROOT/nav.adoc # modules/ai-agents/pages/mcp/index.adoc
Major improvements: - Rename ai-gateway.adoc to gateway-quickstart.adoc for clearer naming - Rename ai-gateway-overview.adoc to gateway-architecture.adoc - Consolidate quickstart-enhanced.adoc into gateway-quickstart.adoc - Remove 161 lines of duplicate content between what-is and architecture pages - Add page metadata (:page-topic-type:, :page-personas:, :learning-objective-N:) to all files - Convert learning objectives from bullets to required attribute format - Create BYOC version requirement partial and add to all 23 AI Gateway pages - Restructure navigation with clearer sections (Overview, Quickstart, Architecture, Observability) - Update all cross-references to renamed files This consolidation reduces content duplication, improves metadata consistency, and provides better content organization for users. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
…docs Updated 13 files to comply with documentation standards for learning objectives: - Added learning objective attributes to metadata (3 files) - Converted all learning objective bullets to checklist format (* [ ]) Files updated: - admin/setup-guide.adoc - builders/connect-your-agent.adoc - builders/discover-gateways.adoc - All 10 integration guide files (admin and user variants for Claude Code, Cline, Continue, Cursor, GitHub Copilot) All 23 AI Gateway files now have proper learning objective format per team standards. Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>
Replace model="auto" with realistic model names in CEL examples, fix UI navigation paths to use Agentic AI sidebar structure, confirm streaming and response format compatibility, uncomment verification command, comment out unverified CEL sections (A/B testing, time-based routing, reject), revise geographic routing to use different providers, and update observability integrations to confirmed Prometheus + OpenTelemetry. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Comment on lines
50
to
57
| Common models to enable: | ||
| + | ||
| -- | ||
| * `openai/gpt-4o` - OpenAI's most capable model | ||
| * `openai/gpt-4o-mini` - Cost-effective OpenAI model | ||
| * `anthropic/claude-sonnet-3.5` - Balanced Anthropic model | ||
| * `anthropic/claude-opus-4` - Anthropic's most capable model | ||
| -- |
There was a problem hiding this comment.
claude-opus-4.6, claude-sonnet-4.5, gpt-5.2, gpt-5.2-mini I think those would be more up-to-date
|
|
||
| *Can be configured:* | ||
|
|
||
| * Provider credentials (OpenAI, Anthropic) |
Update all AI Gateway documentation to use latest model versions: - OpenAI: gpt-5.2, gpt-5.2-mini (was gpt-4o, gpt-4o-mini) - Anthropic: claude-sonnet-4.5, claude-opus-4.6 (was claude-sonnet-3.5, claude-opus-4) - Add Google Gemini: gemini-2.0-flash, gemini-2.0-pro Updated 27 files including core docs, admin/builder guides, integration guides, CEL routing examples, and code samples across all documentation. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Move AI Hub documentation to partials folder to preserve content but prevent rendering in Package 1 release: - Move configure-ai-hub.adoc, eject-to-custom-mode.adoc, use-ai-hub-gateway.adoc, gateway-modes.adoc to partials/ai-hub/ - Wrap all xrefs to AI Hub pages with ifdef::ai-hub-available[] conditionals - Remove AI Hub pages from nav.adoc - Update 7 files with conditional xrefs: gateway-quickstart, setup-guide, discover-gateways, what-is-ai-gateway, gateway-architecture, ai-hub-preference-toggles AI Hub content preserved for future Package 2 release. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Contributor
|
Not seeing a definition for gateway in the ADP glossary. ? |
Feediver1
reviewed
Feb 6, 2026
Feediver1
reviewed
Feb 6, 2026
Feediver1
reviewed
Feb 6, 2026
|
|
||
| * Access to the Redpanda Cloud Console with administrator privileges | ||
| * API keys for at least one LLM provider (OpenAI, Anthropic, Google AI) | ||
| * (Optional) MCP server endpoints if you plan to use tool aggregation |
Feediver1
reviewed
Feb 6, 2026
|
|
||
| Providers represent upstream services (Anthropic, OpenAI, Google AI) and associated credentials. Providers are disabled by default and must be enabled explicitly by an administrator. | ||
|
|
||
| . In the Redpanda Cloud Console, navigate to *AI Gateway* → *Providers*. |
Contributor
There was a problem hiding this comment.
I like the arrow in this nav description. We should standardize on this. Other Cloud docs are using page > page. Add to our UI Style and also our Doc Style guidelines.
Feediver1
reviewed
Feb 6, 2026
| Common models to enable: | ||
| + | ||
| -- | ||
| * `openai/gpt-5.2` - OpenAI's most capable model |
Contributor
There was a problem hiding this comment.
Ooof--this will need to be updated every time one of these versions changes. Need an alert that tells you, otherwise, they will likely fall out of date quickly. Or, does eng need to test and certify each version such that we would recommend in the docs? Just trying to figure out the best way you can stay up-to- date with these. Need a script that tracks the versions on the sites, then updates our doc every time it updates. That would be cool.
Feediver1
reviewed
Feb 6, 2026
Gateway ID is now embedded in the endpoint URL path instead of passed as a separate rp-aigw-id HTTP header. Updated all code examples, UI navigation, and documentation across 22 files. Also updated observability dashboard references to point to the gateway Overview tab where applicable, and commented out references to UI features not yet available. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Update MCP server fields to match Create MCP Server dialog (Server ID, Display Name, Server Address, Defer Loading Override, Forward OIDC Token Override) - Update orchestrator section to reflect system-managed config with configurable blocked tool patterns - Update deferred loading config to use per-server Defer Loading Override dropdown instead of gateway-level toggle - Update observability references to point to gateway Overview tab - Comment out references to UI features not yet available Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Description
This pull request introduces the AI Gateway in Redpanda Cloud.
Resolves https://redpandadata.atlassian.net/browse/DOC-1867
Review deadline: Feb 6
Page previews
Checks