diff --git a/README.md b/README.md index a6ffeb7..4fb5640 100644 --- a/README.md +++ b/README.md @@ -164,6 +164,7 @@ Start here for the deeper deployment details: ## More Resources - [Architecture](./docs/ARCHITECTURE.md) +- [AI-native client matrix](./docs/AI-NATIVE-MATRIX.md) - [API reference](./docs/API.md) - [Configuration reference](./docs/CONFIGURATION.md) - [Operations guide](./docs/OPERATIONS.md) diff --git a/docs/AI-NATIVE-MATRIX.md b/docs/AI-NATIVE-MATRIX.md new file mode 100644 index 0000000..cfe1d66 --- /dev/null +++ b/docs/AI-NATIVE-MATRIX.md @@ -0,0 +1,101 @@ +# FoundryGate AI-Native Client Matrix + +This page tracks the post-`v1.0` client-expansion line for FoundryGate. + +The goal is not to add endless one-off integrations. The goal is to identify which AI-native clients and agent frameworks are worth first-class starter templates, which ones only need a compatibility note, and which ones should wait until there is a real operator need. + +All star and recency figures below were checked against the public GitHub repos on **2026-03-15**. + +## Selection Rules + +We prioritize frameworks that are: + +- actively maintained +- already popular enough to matter externally +- able to use OpenAI-compatible APIs or configurable base URLs cleanly +- likely to benefit from local-worker routing, policy routing, fallback, and operator visibility + +We de-prioritize frameworks that: + +- require deep custom runtime embedding instead of a clean HTTP integration +- are closer to full app platforms than reusable client frameworks +- would force FoundryGate into plugin/API surface commitments too early + +## Priority Matrix + +| Project | Stars | Last push | Fit for FoundryGate | Initial action | +| --- | ---: | --- | --- | --- | +| [langchain-ai/langchain](https://github.com/langchain-ai/langchain) | 129,607 | 2026-03-15 | Very high | Template | +| [microsoft/autogen](https://github.com/microsoft/autogen) | 55,647 | 2026-03-14 | Very high | Template | +| [run-llama/llama_index](https://github.com/run-llama/llama_index) | 47,689 | 2026-03-15 | Very high | Template | +| [crewAIInc/crewAI](https://github.com/crewAIInc/crewAI) | 46,136 | 2026-03-15 | Very high | Template | +| [agno-agi/agno](https://github.com/agno-agi/agno) | 38,706 | 2026-03-15 | High | Template | +| [microsoft/semantic-kernel](https://github.com/microsoft/semantic-kernel) | 27,463 | 2026-03-14 | High | Template | +| [langchain-ai/langgraph](https://github.com/langchain-ai/langgraph) | 26,455 | 2026-03-15 | Very high | Template | +| [deepset-ai/haystack](https://github.com/deepset-ai/haystack) | 24,512 | 2026-03-15 | High | Template | +| [paperclipai/paperclip](https://github.com/paperclipai/paperclip) | 24,080 | 2026-03-15 | High | Requested template | +| [mastra-ai/mastra](https://github.com/mastra-ai/mastra) | 22,021 | 2026-03-15 | High | Template | +| [google/adk-python](https://github.com/google/adk-python) | 18,380 | 2026-03-15 | High | Template | +| [pydantic/pydantic-ai](https://github.com/pydantic/pydantic-ai) | 15,480 | 2026-03-13 | High | Matrix now, template next | +| [camel-ai/camel](https://github.com/camel-ai/camel) | 16,354 | 2026-03-15 | Medium-high | Matrix now, template next | +| [Agent-Field/SWE-AF](https://github.com/Agent-Field/SWE-AF) | 436 | 2026-03-15 | High | Requested template | +| [Heyvhuang/ship-faster](https://github.com/Heyvhuang/ship-faster) | 320 | 2026-03-15 | Medium | Requested template | + +## Matrix-Only For Now + +These projects are important signals, but they are not the best first template targets for `v1.1.0`. + +| Project | Stars | Why not first-wave | +| --- | ---: | --- | +| [langgenius/dify](https://github.com/langgenius/dify) | 132,912 | More platform/application surface than client-template surface | +| [OpenHands/OpenHands](https://github.com/OpenHands/OpenHands) | 69,161 | Strong dev-agent platform, but less “generic starter template” friendly for the first FoundryGate slice | +| [BlockRunAI/ClawRouter](https://github.com/BlockRunAI/ClawRouter) | 5,522 | Best treated as competitive/reference input, not as a target integration | + +## First-Wave Template Set For `v1.1.0` + +This is the recommended minimum useful template wave: + +1. `SWE-AF` +2. `paperclip` +3. `ship-faster` +4. `LangChain` +5. `LangGraph` +6. `AutoGen` +7. `LlamaIndex` +8. `CrewAI` + +That set gives FoundryGate: + +- the three user-requested integrations +- strong coverage of the most recognizable agent ecosystems +- both Python-heavy and orchestration-heavy client shapes +- a good external-discovery story without documenting ten nearly identical setups at once + +## Second-Wave Template Set + +Good candidates once the first wave is stable: + +- `Agno` +- `Semantic Kernel` +- `Haystack` +- `Mastra` +- `Google ADK` +- `PydanticAI` +- `CAMEL` + +## ClawRouter Watch List + +ClawRouter remains a useful reference point for product direction and operator ergonomics. + +We should keep watching for: + +- modality expansion patterns +- clearer comparison/positioning language +- routing-signal ideas worth formalizing better in FoundryGate +- operator-experience improvements that fit the no-build dashboard + +We should not blindly copy: + +- product claims we cannot verify +- hosted-account assumptions that do not fit FoundryGate's local-first shape +- features that would bloat the gateway core instead of strengthening it diff --git a/docs/FOUNDRYGATE-ROADMAP.md b/docs/FOUNDRYGATE-ROADMAP.md index cdbf08e..a5310c1 100644 --- a/docs/FOUNDRYGATE-ROADMAP.md +++ b/docs/FOUNDRYGATE-ROADMAP.md @@ -19,7 +19,21 @@ The foundation that used to be the near-term buildout is largely in place: This roadmap now shifts from "rename and foundation" to "deepen the gateway plane without bloating it". -`v1.0.0` is the current release line: the focus now shifts from feature accretion to a stable gateway baseline, a completed security review gate, and a separate npm CLI surface for Node-facing integrations. +`v1.0.0` is now shipped. The next block should stay disciplined: deepen AI-native client coverage, improve client-facing observability, and refine routing policy without turning FoundryGate into a sprawling platform. + +## Post-1.0 direction + +The first post-`1.0` block should stay narrow enough to ship as `v1.1.0`. + +Primary goals: + +- double-check and extend AI-native client support beyond the current OpenClaw, n8n, and CLI baseline +- ship the next wave of integration starters for requested and high-signal agent frameworks +- expose more useful per-client token and usage metrics in the operator surface +- audit the routing-stage stack so the responsibility of each layer stays clear +- keep a structured watch on ClawRouter-style product evolution without copying features blindly + +The current framework prioritization lives in [AI-NATIVE-MATRIX.md](./AI-NATIVE-MATRIX.md). ## Big Picture @@ -91,6 +105,26 @@ Today the runtime already supports: - route introspection and traces - client-aware routing defaults for OpenClaw, n8n, and CLI callers +The next runtime gap to close is not “more core abstraction”. It is “more real clients with less glue”. + +## `v1.1.0`: AI-native client expansion and operator visibility + +Primary goals: + +- add the first post-`1.0` starter wave for requested and high-signal AI-native clients +- add a curated framework matrix so external users can quickly see where FoundryGate fits +- deepen client and token reporting in API and dashboard surfaces +- review policy, static, heuristic, hook, client-profile, and classifier boundaries with clearer ownership and tests + +Recommended minimal slices: + +1. AI-native client matrix plus roadmap update +2. first-wave starter templates for `SWE-AF`, `paperclip`, `ship-faster`, and the highest-fit external frameworks +3. per-client token and usage reporting in stats and dashboard views +4. routing-layer review plus targeted rule/test cleanup + +The plugin question should stay explicitly out of scope for `v1.1.0` and be revisited only after this release line lands. + ## OpenClaw direction OpenClaw remains a first-class integration surface.