docs: add MockServer, WireMock and agent tool mocking to the homepage comparison

[jpr5]

What changed

• The main comparison table on the homepage has two new columns, MockServer and WireMock. Every existing row has a cell for each of them.
• Two new rows sit directly after "MCP tool mocking": Scenario-scoped MCP tool fakes and Fail on undeclared MCP tool.
• A new Agent tool mocking table sits below the main one. It compares aimock MCP fakes, Mastra Experiment Tool Mocks and MockServer MCP mocking.
• A fix for the homepage scrolling sideways on phones. That bug already existed in the Record section, and a one-line CSS change fixes it. Below 1140px, the tables now scroll sideways inside their own box. From 1140px up, the sticky header works as before.

Only docs/index.html changes.

Evidence for every cell

Each competitor cell is backed by a documented source, checked on 2026-10-05. The aimock cells describe aimock's own features (MCP fakes shipped in #490). The ✗ cells for MSW, VidaiMock, mock-llm, piyook/llm-mock and mokksy/ai-mocks in the two new rows follow from their existing ✗ in "MCP tool mocking".

Main table: MockServer and WireMock cells → source

MockServer pages are under https://www.mock-server.com/mock_server/. WireMock pages are under https://wiremock.org/docs/.

Row	MockServer	Source	WireMock	Source
Cross-process interception	✓	https://www.mock-server.com/	✓	standalone/java-jar/, standalone/docker/, solutions/kubernetes/
Chat Completions SSE	✓	llm_response_mocking.html	Not documented	solutions/ai/
Responses API SSE	✓	llm_response_mocking.html	Not documented	solutions/ai/
Claude Messages API	✓	llm_response_mocking.html	Not documented	solutions/ai/
Gemini streaming	✓	llm_response_mocking.html	Not documented	solutions/ai/
Multi-provider support	✓	llm_response_mocking.html	Not documented	solutions/ai/
OpenRouter router / fallback simulation	Partial (OpenAI-compatible endpoint only; no fallback routing)	llm_response_mocking.html	Not documented	solutions/ai/
WebSocket APIs	Partial (Java client only)	llm_response_mocking.html	Partial (4.x beta, generic WebSocket only)	websockets/
Embeddings API	✓	llm_response_mocking.html	Not documented	solutions/ai/
Image generation	✗	llm_response_mocking.html ("does not ... generate image/audio responses")	Not documented	solutions/ai/
Image editing	✗	llm_response_mocking.html	Not documented	solutions/ai/
Text-to-Speech	✗	llm_response_mocking.html	Not documented	solutions/ai/
Audio transcription	Not documented	all 93 pages in https://www.mock-server.com/sitemap.xml	Not documented	solutions/ai/
Audio translation	Not documented	sitemap scan, as above	Not documented	solutions/ai/
Non-speech audio	✗	llm_response_mocking.html	Not documented	solutions/ai/
Video generation	Not documented	sitemap scan, as above	Not documented	solutions/ai/
AG-UI event mocking	Not documented	sitemap scan, as above	Not documented	all 76 pages in https://wiremock.org/sitemap-0.xml
MCP tool mocking	✓	ai_protocol_mocking.html	Not documented	solutions/ai/
Scenario-scoped MCP tool fakes (new)	Not documented	ai_protocol_mocking.html	Not documented	solutions/ai/
Fail on undeclared MCP tool (new)	Not documented	ai_protocol_mocking.html	Not documented	solutions/ai/
A2A agent mocking	✓	ai_protocol_mocking.html	Not documented	solutions/ai/
Vector DB mocking	Not documented	sitemap scan, as above	Not documented	WireMock sitemap scan, as above
Search & rerank	Partial (rerank only; no web search)	llm_response_mocking.html	Not documented	WireMock sitemap scan, as above
Record & replay	✓	ai_traffic_inspection.html, configuration_properties.html	✓ HTTP	record-playback/
Drift detection	Partial (HTTP stubs; LLM cassette drift named, not described)	drift_detection.html	✗ OSS	wiremock-cloud/ (OSS vs Cloud table)
Sequential / stateful responses	✓	llm_response_mocking.html	✓ HTTP	stateful-behaviour/
Request journal	✓	debugging_issues.html	✓	verifying/
Error injection	✓	https://www.mock-server.com/	✓ HTTP	simulating-faults/
Chaos testing	✓	https://www.mock-server.com/	✗ OSS	wiremock-cloud/ (OSS vs Cloud table)
Streaming physics	✓	llm_response_mocking.html	Partial (chunked dribble delay only)	simulating-faults/
Structured output / JSON mode	✓	creating_expectations.html	Not documented	solutions/ai/
Programmatic API	✓ 8 languages	https://www.mock-server.com/	✓ Java/JVM	java-usage/
Fixture files	✓ JSON	initializing_expectations.html	✓ JSON	standalone/java-jar/
GitHub Action	✓	https://github.com/marketplace/actions/setup-mockserver	Not documented	WireMock sitemap scan, as above
Vitest / Jest plugins	Partial (framework-agnostic helper, hooks wired by hand)	mockserver_clients.html	Not documented	WireMock sitemap scan, as above
Docker + Helm	✓	https://www.mock-server.com/	✓ (Helm experimental)	standalone/docker/, solutions/kubernetes/
Prometheus metrics	✓	https://www.mock-server.com/	Partial (community extension only)	external-resources/
Streaming usage chunks	Partial (realtime WebSocket only)	llm_response_mocking.html	Not documented	solutions/ai/
Rate limiting headers	Partial (generic 429 rate-limit clause only)	llm_response_mocking.html, configuration_properties.html	Not documented	solutions/ai/
Dependencies	JVM, or JVM-less binary	https://www.mock-server.com/	JVM	standalone/java-jar/

Agent tool mocking table: Mastra and MockServer cells → source

Mastra sources: docs https://mastra.ai/docs/evals/experiments and blog https://mastra.ai/blog/introducing-experiment-tool-mocks. MockServer sources are under https://www.mock-server.com/mock_server/.

Row	Mastra	Source	MockServer	Source
Works with any agent framework	✗ (agent targets only)	docs	✓ (over MCP HTTP)	ai_protocol_mocking.html
MCP tools	Not documented	docs, blog	✓	ai_protocol_mocking.html
In-process (non-network) tools	✓	docs	n/a (a network mock cannot see in-process calls)	—
Ordered answers per tool	✓	docs	Not documented	ai_protocol_mocking.html
Exact-argument matching	✓	docs	Not documented	ai_protocol_mocking.html
Any-argument matching	✓	docs	✓ (one answer for every call)	ai_protocol_mocking.html
Fail on undeclared tool	✓	docs	Not documented	ai_protocol_mocking.html
Scripted tool errors	Not documented	docs	✓	ai_protocol_mocking.html
Per-test state and reset	✓	docs	Not documented	ai_protocol_mocking.html
LLM responses mocked in the same scenario	Not documented	docs, blog	Partial (same server; tool mocks not tied to the LLM scenario)	llm_response_mocking.html, ai_protocol_mocking.html
Create fakes from recorded runs	Partial (how a mock is made from a trace is not described)	docs	✓	ai_traffic_inspection.html, ai_mcp_tools.html
Report unused mocks	✓	docs	Not documented	ai_protocol_mocking.html
Workflows	Planned	blog	n/a	—

Proof

• Competitor scan dry run. Before: "Parsed 38 capability rows, 6 competitor columns". After: "Parsed 40 capability rows, 8 competitor columns".
• Tests. The competitive-matrix test suite passes against the new page. No test was edited.
• Live demo. A live demo with the real MCP client shows that both new rows ("Scenario-scoped MCP tool fakes" and "Fail on undeclared MCP tool") are true for aimock.
• Screenshots. Checked at 375, 430, 900, 1024, 1280 and 1440px wide. At 375 and 430 the page no longer scrolls sideways; the tables scroll inside their own box and the rest of the page stays put. At 900 and 1024 the same happens, with the table scrollbar at the bottom of the table. At 1280 and 1440 the full table fits with all nine columns visible and the header stays sticky while scrolling. The page never scrolls sideways at any width.

Known limits and follow-ups

• From 1140px up the table fits with only about 10px to spare. That is the price of keeping the sticky header, so a future wider column needs another look.
• Between 769 and 1139px the sideways scrollbar sits at the bottom of a tall table.
• The weekly competitor scan does not track the MockServer and WireMock columns yet. A follow-up PR handles that.
• The scan does not update provider-count cells on migration pages (replaceProviderCount). This predates this PR.

Release note

Merging deploys the homepage, which then advertises MCP fakes. They are on main (#490) but not in any npm release yet. Consider releasing first, or merging alongside a release.

🤖 Generated with Claude Code

https://claude.ai/code/session_01AKuRWApa4KRBia1X1F3anz

[pkg-pr-new]

Open in StackBlitz

npm i https://pkg.pr.new/@copilotkit/aimock@494

commit: d0398c0