Sprint 11: benchmark_v3.py — robustness + competitor-style benchmark suite

Sprint 10C: lora_plan.py + distill.py — adapter/distillation planning (dry-run, no GPU required)

Sprint 10B: optimizer.py — agentic decision policy (improving/plateau/degrading/rollback)

Sprint 10A: shadow_eval.py — compare candidate vs baseline safely

Sprint 9C: prompt_pack.py — epigenetic optimization (prompts before weights)

Sprint 9B: dataset.py — trace → filtered training dataset builder

Sprint 9A: fingerprint.py — capability fingerprint from execution traces

Sprint 9-10: optimization package init

Sprint 7: skills/ci.py — skill testing, promotion, rollback, immune scan

Sprint 7: skills/schema.py — SkillCard + SkillGenome with versioning

Sprint 7: skills package — SkillCard schema, runtime, CI, export

Sprint 6: mas_generator.py — use-case → generated multi-agent system

Sprint 5: routing.py — LLMCallRouter, ModelSelector, cost homeostasis

Sprint 8: quorum.py — consensus/disagreement topology switches + critic ensemble

Sprint 4D: AGENTS.md parser

Sprint 4C: AG-UI adapter — PAEvent to frontend stream + SSE helper

Sprint 4B: A2A — AgentCard, A2AClient for agent-to-agent delegation

Sprint 4A: MCP bridge — discover, validate, call remote tools via MCP

Sprint 4: protocols package init

Sprint 3: memory_homeostasis.py — budget, archive, consolidation, Q-retriever

Sprint 2: update runtime __init__ with checkpoint + state exports

Sprint 2: checkpoint.py — Checkpointer protocol + InMemory/JSONL/SQLite implementations

Sprint 2: runtime/state.py — typed RunState for durable execution

v3.0.0a1: add runtime event system exports, bump version

Sprint 1: streaming_v3 — AG-UI compatible adapter + legacy bridge

Sprint 1: EventBus — async pub/sub with backpressure, replay, lane isolation

Sprint 1: canonical event schema (PAEvent, EventKind, Visibility)

Sprint 1: runtime package init

v2.1.1: sync version with PyPI

v2.1.0: creative names (Spark/Flow/swarm/Council/Vault) + prod test 19/19 pass

v2.1.0: creative names (Spark/Flow/swarm/Council/Vault) + prod test 19/19 pass

Track 4: breakthroughs + TOML prompts — purpose_agent/breakthroughs.py

Track 3: TOML prompts + PURPOSE_LEARNING.md whitepaper — purpose_agent/purpose_function.py

Track 3: TOML prompts + PURPOSE_LEARNING.md whitepaper — purpose_agent/actor.py

Track 3: TOML prompts + PURPOSE_LEARNING.md whitepaper — purpose_agent/robust_parser.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/orchestrator.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/optimizer.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/purpose_function.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/actor.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/llm_backend.py

fix: universal parsing + OpenRouter + state bug — purpose_agent/robust_parser.py

fix: real-model robustness — purpose_agent/purpose_function.py

fix: real-model robustness — purpose_agent/actor.py

Track 1: purpose_agent/slm_backends.py

Track 1: purpose_agent/experience_replay.py

Track 1: purpose_agent/tools.py

Track 1: purpose_agent/llm_backend.py

Track 1: purpose_agent/__init__.py

release: Purpose Agent v2.0.0 — updated __init__.py with all V2 + research exports

V2 merge: purpose_agent/unified.py