9 projects
phionyx-eval
Phionyx evaluation tooling — an LLM-as-judge primitive (eval-side) that composes with AIREP (the AI Runtime Evidence Protocol) decision records. Producers run a judge against a (claim, evidence, rubric) triple and emit a signed Judgment envelope. Caller supplies the LLM client; no hard dependency on any provider SDK.
phionyx-openai-agents
OpenAI Agents SDK tracing bridge for Phionyx runtime evidence — every Trace and Span emitted by the SDK becomes a hash-chained envelope entry (signed when a signer is configured; unsigned by default).
phionyx-langchain-langgraph
Native LangChain + LangGraph adapters for Phionyx runtime evidence — every chain, tool call, and LLM invocation emits a hash-chained envelope entry (signed when a signer is configured; unsigned by default).
phionyx-letta
Letta memory-mutation adapter for Phionyx runtime evidence — every memory write, clear, forget, or consolidation emits a hash-chained envelope (signed when a signer is configured) with before/after diff.
phionyx-mcp-server
Phionyx MCP trust boundary governance layer — descriptor signing, runtime guardrail, tamper-evident audit chain for MCP-capable hosts.
phionyx-pipeline-mcp
Phionyx self-governance gate for Claude Code — three-layer verification (LLM declaration → git diff truth → deterministic gate) over the agent's own 'fixed / tested / changed' claims. v0.3.0 adds opt-in evidence binding: the require_tool directive (a factual claim without same-turn evidence must bind a tool first) and continuity binding (catches 'read-but-not-bound'). Both default-OFF / non-regressive; detector calibration is not yet validated.
phionyx-core
Phionyx Core SDK - AI↔AI Mediator and Echo Engine
phionyx-compliance
Phionyx compliance evidence report draft generator — Article-13-oriented and other governance-framework mappings from RGE v0.2 audit chains.
phionyx-eval-inspect
Interoperability bridge — export Phionyx Reasoned Governance Envelope (RGE) evidence into Inspect AI evaluation logs.
