langgraph-macp 0.1.0__py3-none-any.whl

langgraph_macp/__init__.py

@@ -0,0 +1,39 @@
+"""LangGraph integration for MACP: add a binding multi-agent decision to a graph."""
+
+from __future__ import annotations
+
+from importlib.metadata import version as _version
+
+from .llm import DEFAULT_SYSTEM_PROMPT, UNPARSEABLE, llm_evaluator, parse_evaluation
+from .node import (
+    AUTHORITY_SCOPE,
+    DECISION_PROPOSAL_ID,
+    DEFAULT_ACTION,
+    DEFAULT_TTL_MS,
+    QUORUM_REQUEST_ID,
+    ConsensusParticipant,
+    ConsensusResult,
+    MacpConsensusNode,
+    consensus_receipt,
+    normalize_recommendation,
+)
+
+__version__ = _version("langgraph-macp")
+
+__all__ = [
+    "AUTHORITY_SCOPE",
+    "DECISION_PROPOSAL_ID",
+    "DEFAULT_ACTION",
+    "DEFAULT_SYSTEM_PROMPT",
+    "DEFAULT_TTL_MS",
+    "QUORUM_REQUEST_ID",
+    "UNPARSEABLE",
+    "ConsensusParticipant",
+    "ConsensusResult",
+    "MacpConsensusNode",
+    "__version__",
+    "consensus_receipt",
+    "llm_evaluator",
+    "normalize_recommendation",
+    "parse_evaluation",
+]

langgraph_macp/llm.py

@@ -0,0 +1,89 @@
+"""``llm_evaluator``: a LangChain chat model as an MACP ``EvaluationStrategy``."""
+
+from __future__ import annotations
+
+import json
+import math
+import re
+from typing import Any
+
+from langchain_core.language_models import BaseChatModel
+from langchain_core.messages import HumanMessage, SystemMessage
+from macp_sdk.agent import EvaluationResult, SessionInfo
+
+from .node import normalize_recommendation
+
+DEFAULT_SYSTEM_PROMPT = (
+    "You are one participant in a binding multi-agent decision. Evaluate the proposal "
+    "you are given. Answer with a single JSON object and nothing else: "
+    '{"recommendation": "APPROVE" | "REJECT" | "REVIEW" | "BLOCK", '
+    '"confidence": <number from 0 to 1>, "reason": "<one sentence>"}.'
+)
+UNPARSEABLE = "unparseable model output"
+
+_FENCE = re.compile(r"^```[a-zA-Z0-9_-]*\s*\n?(.*?)\n?```$", re.DOTALL)
+
+
+def parse_evaluation(text: str) -> EvaluationResult:
+    """Parse a model's JSON answer. Any defect gives ``REVIEW`` with confidence 0."""
+    body = text.strip()
+    fenced = _FENCE.match(body)
+    if fenced:
+        body = fenced.group(1).strip()
+    try:
+        data = json.loads(body)
+        if not isinstance(data, dict):
+            raise ValueError("not an object")
+        recommendation = normalize_recommendation(data.get("recommendation"))
+        confidence = data.get("confidence")
+        if (
+            isinstance(confidence, bool)
+            or not isinstance(confidence, int | float)
+            or not math.isfinite(confidence)
+            or not 0 <= confidence <= 1
+        ):
+            raise ValueError("confidence out of range")
+        reason = data.get("reason", "")
+        return EvaluationResult(
+            recommendation=recommendation,
+            confidence=float(confidence),
+            reason=reason if isinstance(reason, str) else json.dumps(reason),
+        )
+    except ValueError:
+        return EvaluationResult(recommendation="REVIEW", confidence=0.0, reason=UNPARSEABLE)
+
+
+class _LlmEvaluator:
+    __slots__ = ("_model", "_system_prompt")
+
+    def __init__(self, model: BaseChatModel, system_prompt: str) -> None:
+        self._model = model
+        self._system_prompt = system_prompt
+
+    def evaluate(self, proposal: dict[str, Any], context: SessionInfo) -> EvaluationResult:
+        question = {
+            "session_id": context.session_id,
+            "mode": proposal["mode"],
+            "option": proposal["option"],
+            "rationale": proposal["rationale"],
+            "details": proposal["details"],
+        }
+        answer = self._model.invoke(
+            [
+                SystemMessage(content=self._system_prompt),
+                HumanMessage(content=json.dumps(question, default=str)),
+            ]
+        )
+        return parse_evaluation(answer.text)
+
+
+def llm_evaluator(model: BaseChatModel, *, system_prompt: str | None = None) -> _LlmEvaluator:
+    """Wrap a LangChain chat model as an ``EvaluationStrategy``.
+
+    The model is asked for ``{recommendation, confidence, reason}`` as JSON.
+    A Markdown code fence around the answer is fine. ``ABSTAIN`` becomes
+    ``REVIEW``. An answer that cannot be read becomes ``REVIEW`` with
+    confidence 0 (reason ``"unparseable model output"``), so a bad answer
+    never sinks the session.
+    """
+    return _LlmEvaluator(model, system_prompt or DEFAULT_SYSTEM_PROMPT)

langgraph_macp/node.py

@@ -0,0 +1,434 @@
+"""``MacpConsensusNode``: a LangGraph node that settles a binding MACP decision.
+
+The node's client is the session's initiator: it opens the session, puts the
+proposal, and commits. It never votes. Each participant evaluates through an
+SDK ``EvaluationStrategy`` and casts its ballot under its own auth:
+
+- **Decision** (``macp.mode.decision.v1``): the runtime only accepts a
+  ``Proposal`` from a declared participant, so the initiator is listed first.
+  It sends ``Proposal`` ``p1``. Every participant sends an ``Evaluation``,
+  then they vote in list order.
+- **Quorum** (``macp.mode.quorum.v1``): the initiator is not a participant.
+  It sends ``ApprovalRequest`` ``r1``, and each participant answers with
+  ``Approve`` / ``Reject`` / ``Abstain``.
+
+The outcome is approved once ``threshold`` participants approve. It is
+rejected as soon as the approvals plus the participants yet to vote fall
+short of the threshold. The initiator commits that outcome at once, and the
+remaining participants never vote. The Commitment, its hash and the tally
+are written to graph state under ``output_key``.
+"""
+
+from __future__ import annotations
+
+import asyncio
+import contextlib
+import json
+from collections.abc import Callable, Mapping, Sequence
+from dataclasses import dataclass
+from datetime import UTC, datetime
+from typing import Any, Literal, TypedDict
+
+from langchain_core.runnables import RunnableLambda
+from macp_sdk import (
+    DEFAULT_POLICY_VERSION,
+    MODE_DECISION,
+    MODE_QUORUM,
+    AuthConfig,
+    DecisionReceipt,
+    DecisionSession,
+    MacpClient,
+    QuorumSession,
+    ReceiptSigner,
+    build_receipt,
+    commitment_from_projection,
+    commitment_hash,
+)
+from macp_sdk.agent import EvaluationResult, EvaluationStrategy, SessionInfo, VotingStrategy
+from macp_sdk.commitment_hash import canonical_projection
+from macp_sdk.validation import validate_confidence, validate_vote
+
+Mode = Literal["decision", "quorum"]
+Outcome = Literal["approve", "reject"]
+
+DECISION_PROPOSAL_ID = "p1"
+QUORUM_REQUEST_ID = "r1"
+DEFAULT_ACTION = "macp.consensus"
+AUTHORITY_SCOPE = "macp.consensus"
+DEFAULT_TTL_MS = 60_000
+
+_MODES: Mapping[str, str] = {"decision": MODE_DECISION, "quorum": MODE_QUORUM}
+
+#: A wire-valid Decision ``Evaluation`` recommendation, as the runtime accepts it.
+_RECOMMENDATIONS = frozenset({"APPROVE", "REVIEW", "BLOCK", "REJECT"})
+#: Recommendation → Decision vote. Quorum uses the same map, title-cased.
+_BALLOTS: Mapping[str, str] = {
+    "APPROVE": "APPROVE",
+    "REVIEW": "ABSTAIN",
+    "REJECT": "REJECT",
+    "BLOCK": "REJECT",
+}
+
+EvaluatorFn = Callable[[dict[str, Any], SessionInfo], "EvaluationResult | Mapping[str, Any]"]
+Evaluator = EvaluationStrategy | EvaluatorFn
+
+
+@dataclass(frozen=True)
+class ConsensusParticipant:
+    """One voter: its agent id, how it evaluates, and how it authenticates.
+
+    ``auth`` defaults to dev-agent auth for ``id``; pass a bearer
+    ``AuthConfig`` in production. ``voter`` (Decision only) overrides the
+    default vote, which maps the participant's own recommendation.
+    """
+
+    id: str
+    evaluator: Evaluator
+    auth: AuthConfig | None = None
+    voter: VotingStrategy | None = None
+
+
+class ConsensusResult(TypedDict):
+    """What the node writes under ``output_key``.
+
+    ``commitment`` is the RFC-MACP-0013 §3 projection of the committed
+    ``CommitmentPayload`` (a plain dict, so checkpointers can store it).
+    """
+
+    outcome: Outcome
+    session_id: str
+    commitment_hash: str
+    mode: str
+    commitment: dict[str, Any]
+    approvals: int
+    threshold: int
+
+
+def normalize_recommendation(value: object) -> str:
+    """Uppercase ``value`` and check it is APPROVE / REVIEW / BLOCK / REJECT.
+
+    ``ABSTAIN`` becomes ``REVIEW``, so every result is a wire-valid Decision
+    ``Evaluation``. Anything else raises ``ValueError``.
+    """
+    if not isinstance(value, str):
+        raise ValueError(f"recommendation must be a string, got {type(value).__name__}")
+    normalized = value.strip().upper()
+    if normalized == "ABSTAIN":
+        return "REVIEW"
+    if normalized not in _RECOMMENDATIONS:
+        raise ValueError(
+            f"invalid recommendation {value!r}: must be one of APPROVE, REVIEW, BLOCK, REJECT"
+        )
+    return normalized
+
+
+def _as_result(value: EvaluationResult | Mapping[str, Any]) -> EvaluationResult:
+    if isinstance(value, Mapping):
+        value = EvaluationResult(
+            recommendation=value.get("recommendation"),  # type: ignore[arg-type]
+            confidence=value.get("confidence"),  # type: ignore[arg-type]
+            reason=str(value.get("reason", "")),
+        )
+    if not isinstance(value, EvaluationResult):
+        raise TypeError(
+            f"evaluator returned {type(value).__name__}; expected EvaluationResult or a mapping"
+        )
+    confidence = value.confidence
+    if isinstance(confidence, bool) or not isinstance(confidence, int | float):
+        raise ValueError(f"confidence must be a number, got {confidence!r}")
+    validate_confidence(float(confidence))
+    return EvaluationResult(
+        recommendation=normalize_recommendation(value.recommendation),
+        confidence=float(confidence),
+        reason=str(value.reason),
+    )
+
+
+def _evaluate(
+    evaluator: Evaluator, proposal: dict[str, Any], info: SessionInfo
+) -> EvaluationResult:
+    evaluate = getattr(evaluator, "evaluate", None)
+    raw = evaluate(proposal, info) if callable(evaluate) else evaluator(proposal, info)  # type: ignore[operator]
+    return _as_result(raw)
+
+
+def _default_proposal(state: Mapping[str, Any]) -> Mapping[str, Any] | str:
+    if "proposal" not in state:
+        raise KeyError("MacpConsensusNode: state has no 'proposal' (or pass proposal=)")
+    return state["proposal"]  # type: ignore[no-any-return]
+
+
+def _read_proposal(value: Mapping[str, Any] | str) -> tuple[str, str, Any]:
+    option: object
+    rationale: object
+    if isinstance(value, str):
+        option, rationale, details = value, "", None
+    elif isinstance(value, Mapping):
+        option = value.get("option")
+        rationale = value.get("rationale") or ""
+        details = value.get("details")
+    else:
+        raise TypeError(f"proposal must be a string or a mapping, got {type(value).__name__}")
+    if not isinstance(option, str) or not option:
+        raise ValueError("proposal option must be a non-empty string")
+    if not isinstance(rationale, str):
+        raise ValueError("proposal rationale must be a string")
+    return option, rationale, details
+
+
+def _initiator_of(client: MacpClient, initiator: str | None) -> str:
+    agent = initiator or (client.auth.sender_hint if client.auth else None)
+    if not agent:
+        raise ValueError(
+            "MacpConsensusNode: cannot tell the initiator's agent id from the client's auth; "
+            "pass initiator="
+        )
+    return agent
+
+
+class _Tally:
+    def __init__(self, voters: int, threshold: int) -> None:
+        self.voters = voters
+        self.threshold = threshold
+        self.cast = 0
+        self.approvals = 0
+
+    def add(self, approved: bool) -> None:
+        self.cast += 1
+        self.approvals += approved
+
+    def outcome(self) -> Outcome | None:
+        if self.approvals >= self.threshold:
+            return "approve"
+        if self.approvals + (self.voters - self.cast) < self.threshold:
+            return "reject"
+        return None
+
+
+class MacpConsensusNode(RunnableLambda[Any, dict[str, ConsensusResult]]):
+    """A LangGraph node that settles one MACP Decision or Quorum session.
+
+    Add it like any node: ``graph.add_node("consensus", MacpConsensusNode(...))``.
+    It works under both ``invoke`` and ``ainvoke``. The async path runs the
+    blocking gRPC calls in a worker thread.
+
+    Args:
+        client: A ``MacpClient`` authenticated as the initiator.
+        participants: The voters, in the order they evaluate and vote.
+        mode: ``"decision"`` (default) or ``"quorum"``.
+        policy: A registered ``policy_version`` (default ``policy.default``).
+        threshold: Approvals required (default a strict majority, ``n // 2 + 1``).
+        proposal: ``state -> str | {option, rationale?, details?}``. The
+            default reads ``state["proposal"]``.
+        output_key: The state key written (default ``"macp"``). Declare it in
+            your state schema, or LangGraph drops it.
+        ttl_ms: Session TTL (default 60 s).
+        action: The commitment action stem, committed as ``<action>.approved``
+            or ``<action>.rejected``.
+        initiator: The initiator's agent id, when the client's auth does not
+            carry it (a production bearer token).
+    """
+
+    def __init__(
+        self,
+        client: MacpClient,
+        participants: Sequence[ConsensusParticipant],
+        mode: Mode = "decision",
+        policy: str | None = None,
+        *,
+        threshold: int | None = None,
+        proposal: Callable[[Mapping[str, Any]], Mapping[str, Any] | str] | None = None,
+        output_key: str = "macp",
+        ttl_ms: int = DEFAULT_TTL_MS,
+        action: str = DEFAULT_ACTION,
+        initiator: str | None = None,
+        name: str = "macp_consensus",
+    ) -> None:
+        if mode not in _MODES:
+            raise ValueError(f"mode must be 'decision' or 'quorum', got {mode!r}")
+        voters = list(participants)
+        if not voters:
+            raise ValueError("MacpConsensusNode needs at least one participant")
+        ids = [p.id for p in voters]
+        if len(set(ids)) != len(ids):
+            raise ValueError(f"participant ids must be unique: {ids}")
+        self._initiator = _initiator_of(client, initiator)
+        if self._initiator in ids:
+            raise ValueError(
+                f"the initiator {self._initiator!r} commits and cannot also be a participant"
+            )
+        n = len(voters)
+        self._threshold = n // 2 + 1 if threshold is None else threshold
+        if isinstance(self._threshold, bool) or not 1 <= self._threshold <= n:
+            raise ValueError(f"threshold must be between 1 and {n}, got {threshold!r}")
+        if ttl_ms <= 0:
+            raise ValueError("ttl_ms must be positive")
+        self._client = client
+        self._participants = voters
+        self._mode = mode
+        self._policy = policy or DEFAULT_POLICY_VERSION
+        self._proposal = proposal or _default_proposal
+        self._output_key = output_key
+        self._ttl_ms = ttl_ms
+        self._action = action
+        super().__init__(self._run, afunc=self._arun, name=name)
+
+    async def _arun(self, state: Mapping[str, Any]) -> dict[str, ConsensusResult]:
+        return await asyncio.to_thread(self._run, state)
+
+    def _run(self, state: Mapping[str, Any]) -> dict[str, ConsensusResult]:
+        option, rationale, details = _read_proposal(self._proposal(state))
+        ids = [p.id for p in self._participants]
+        session: DecisionSession | QuorumSession
+        if self._mode == "decision":
+            session = DecisionSession(self._client, policy_version=self._policy)
+            members = [self._initiator, *ids]
+            proposal_id = DECISION_PROPOSAL_ID
+        else:
+            session = QuorumSession(self._client, policy_version=self._policy)
+            members = ids
+            proposal_id = QUORUM_REQUEST_ID
+        session.start(intent=option, participants=members, ttl_ms=self._ttl_ms)
+        try:
+            tally = self._settle(session, proposal_id, option, rationale, details, members)
+            outcome = tally.outcome() or "reject"
+            session.commit(
+                action=f"{self._action}.{'approved' if outcome == 'approve' else 'rejected'}",
+                authority_scope=AUTHORITY_SCOPE,
+                reason=f"{tally.approvals}/{tally.voters} approvals (threshold {tally.threshold})",
+                outcome_positive=outcome == "approve",
+            )
+        except BaseException:
+            with contextlib.suppress(Exception):
+                session.cancel(reason="MacpConsensusNode failed")
+            raise
+        payload = session.projection.commitment
+        if payload is None:  # pragma: no cover - an acked commit is always projected
+            raise RuntimeError("the runtime acked the Commitment but it was not projected")
+        result: ConsensusResult = {
+            "outcome": outcome,
+            "session_id": session.session_id,
+            "commitment_hash": commitment_hash(payload),
+            "mode": _MODES[self._mode],
+            "commitment": json.loads(canonical_projection(payload)),
+            "approvals": tally.approvals,
+            "threshold": tally.threshold,
+        }
+        return {self._output_key: result}
+
+    def _settle(
+        self,
+        session: DecisionSession | QuorumSession,
+        proposal_id: str,
+        option: str,
+        rationale: str,
+        details: Any,
+        members: list[str],
+    ) -> _Tally:
+        wire_details = (
+            b"" if details is None else json.dumps(details, separators=(",", ":")).encode()
+        )
+        if isinstance(session, DecisionSession):
+            session.propose(proposal_id, option, rationale=rationale, supporting_data=wire_details)
+        else:
+            session.request_approval(
+                proposal_id,
+                option,
+                summary=rationale,
+                details=wire_details,
+                required_approvals=self._threshold,
+            )
+        mode = _MODES[self._mode]
+        proposal = {
+            "proposal_id": proposal_id,
+            "option": option,
+            "rationale": rationale,
+            "details": details,
+            "mode": mode,
+        }
+        info = SessionInfo(
+            session_id=session.session_id,
+            mode=mode,
+            participants=list(members),
+            mode_version=session.mode_version,
+            configuration_version=session.configuration_version,
+            policy_version=session.policy_version,
+        )
+        tally = _Tally(len(self._participants), self._threshold)
+        if isinstance(session, DecisionSession):
+            results = []
+            for p in self._participants:
+                result = _evaluate(p.evaluator, proposal, info)
+                session.evaluate(
+                    proposal_id,
+                    result.recommendation,
+                    confidence=result.confidence,
+                    reason=result.reason,
+                    sender=p.id,
+                    auth=self._auth(p),
+                )
+                results.append(result)
+            for p, result in zip(self._participants, results, strict=True):
+                vote, reason = self._vote(session, p, result)
+                if vote is not None:
+                    session.vote(proposal_id, vote, reason=reason, sender=p.id, auth=self._auth(p))
+                    tally.add(vote == "APPROVE")
+                else:
+                    tally.add(False)  # declined to vote: no approval from it
+                if tally.outcome() is not None:
+                    break
+        else:
+            for p in self._participants:
+                result = _evaluate(p.evaluator, proposal, info)
+                ballot = _BALLOTS[result.recommendation]
+                send = {"APPROVE": session.approve, "REJECT": session.reject}.get(
+                    ballot, session.abstain
+                )
+                send(proposal_id, reason=result.reason, sender=p.id, auth=self._auth(p))
+                tally.add(ballot == "APPROVE")
+                if tally.outcome() is not None:
+                    break
+        return tally
+
+    @staticmethod
+    def _vote(
+        session: DecisionSession, p: ConsensusParticipant, result: EvaluationResult
+    ) -> tuple[str | None, str]:
+        if p.voter is None:
+            return _BALLOTS[result.recommendation], result.reason
+        projection = session.decision_projection
+        if not p.voter.should_vote(projection):
+            return None, ""
+        decision = p.voter.decide_vote(projection)
+        return validate_vote(decision.vote), decision.reason
+
+    @staticmethod
+    def _auth(p: ConsensusParticipant) -> AuthConfig:
+        return p.auth or AuthConfig.for_dev_agent(p.id)
+
+
+def _now() -> str:
+    return datetime.now(UTC).strftime("%Y-%m-%dT%H:%M:%SZ")
+
+
+def consensus_receipt(
+    result: ConsensusResult,
+    *,
+    issuer: str,
+    issued_at: str | None = None,
+    signer: ReceiptSigner | None = None,
+) -> DecisionReceipt:
+    """A ``macp-receipt/1`` decision receipt for the node's result, in one call.
+
+    ``issued_at`` defaults to now (``YYYY-MM-DDTHH:MM:SSZ``). Pass a
+    ``ReceiptSigner`` to add an Ed25519 attestation. Without one, the
+    receipt verifies as ``integrity_only``.
+    """
+    return build_receipt(
+        commitment_from_projection(result["commitment"]),
+        session_id=result["session_id"],
+        mode=result["mode"],
+        issuer=issuer,
+        issued_at=issued_at or _now(),
+        signer=signer,
+    )

langgraph_macp-0.1.0.dist-info/METADATA

@@ -0,0 +1,243 @@
+Metadata-Version: 2.4
+Name: langgraph-macp
+Version: 0.1.0
+Summary: LangGraph node that settles a binding MACP Decision or Quorum session
+Author-email: Multi-Agent Coordination Protocol <macp@multiagentcoordinationprotocol.org>
+License: Apache-2.0
+Project-URL: Homepage, https://multiagentcoordinationprotocol.io
+Project-URL: Repository, https://github.com/multiagentcoordinationprotocol/macp-sdk-python
+Project-URL: Source, https://github.com/multiagentcoordinationprotocol/macp-sdk-python/tree/main/integrations/langgraph
+Project-URL: Changelog, https://github.com/multiagentcoordinationprotocol/macp-sdk-python/blob/main/integrations/langgraph/CHANGELOG.md
+Project-URL: Bug Tracker, https://github.com/multiagentcoordinationprotocol/macp-sdk-python/issues
+Keywords: langgraph,langchain,macp,multi-agent,consensus,coordination
+Classifier: Development Status :: 3 - Alpha
+Classifier: Intended Audience :: Developers
+Classifier: License :: OSI Approved :: Apache Software License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Software Development :: Libraries
+Requires-Python: >=3.11
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Requires-Dist: macp-sdk-python[receipts]<0.20,>=0.19.0
+Requires-Dist: langgraph<2,>=1.0
+Requires-Dist: langchain-core<2,>=1.0
+Provides-Extra: examples
+Requires-Dist: langchain-anthropic<2,>=1.0; extra == "examples"
+Provides-Extra: dev
+Requires-Dist: pytest>=8.0; extra == "dev"
+Requires-Dist: pytest-cov>=5.0; extra == "dev"
+Requires-Dist: mypy>=1.10; extra == "dev"
+Requires-Dist: ruff>=0.5; extra == "dev"
+Requires-Dist: build>=1.0; extra == "dev"
+Requires-Dist: twine>=5.0; extra == "dev"
+Dynamic: license-file
+
+# langgraph-macp
+
+Add a **binding multi-agent decision** to a LangGraph graph in a few lines.
+
+`MacpConsensusNode` is a LangGraph node. It opens a [MACP](https://multiagentcoordinationprotocol.io)
+Decision or Quorum session on the MACP runtime and lets each participant evaluate the proposal,
+usually through its own LLM. It then has them vote, commits the outcome, and writes the
+Commitment back into graph state. Anyone can check the committed outcome later: the runtime
+records it, `commitment_hash` identifies it, and a decision receipt for it takes one call.
+
+```bash
+pip install langgraph-macp
+```
+
+The TypeScript twin is
+[`@multiagentcoordinationprotocol/langgraph`](https://github.com/multiagentcoordinationprotocol/macp-sdk-typescript/tree/main/integrations/langgraph).
+It has the same options, the same flow and the same state keys, in camelCase.
+
+## Quick start
+
+```python
+from typing import TypedDict
+
+from langchain_anthropic import ChatAnthropic
+from langgraph.graph import END, START, StateGraph
+from macp_sdk import AuthConfig, MacpClient
+
+from langgraph_macp import (
+    ConsensusParticipant, ConsensusResult, MacpConsensusNode, consensus_receipt, llm_evaluator,
+)
+
+
+class State(TypedDict, total=False):
+    proposal: str
+    macp: ConsensusResult  # declare the output key, or LangGraph drops it
+
+
+client = MacpClient(target="127.0.0.1:50051", allow_insecure=True,
+                    auth=AuthConfig.for_dev_agent("coordinator"))
+model = ChatAnthropic(model="claude-opus-5-5")
+consensus = MacpConsensusNode(client, [
+    ConsensusParticipant("security", llm_evaluator(model, system_prompt="You review security...")),
+    ConsensusParticipant("sre", llm_evaluator(model, system_prompt="You review reliability...")),
+    ConsensusParticipant("product", llm_evaluator(model, system_prompt="You review user impact...")),
+])
+
+graph = StateGraph(State)
+graph.add_node("consensus", consensus)
+graph.add_edge(START, "consensus")
+graph.add_edge("consensus", END)
+
+final = graph.compile().invoke({"proposal": "ship v2 to all users"})
+final["macp"]["outcome"]          # "approve" | "reject"
+final["macp"]["commitment_hash"]  # "sha256:…"
+receipt = consensus_receipt(final["macp"], issuer="did:web:example.com")
+```
+
+A local runtime for development:
+
+```bash
+docker run -d -p 50051:50051 -e MACP_ALLOW_INSECURE=1 \
+  ghcr.io/multiagentcoordinationprotocol/macp-runtime:0.9.0
+```
+
+## How the node runs a session
+
+```mermaid
+sequenceDiagram
+    participant G as Graph (initiator)
+    participant R as MACP runtime
+    participant A as participant 1..n
+    G->>R: SessionStart (participants)
+    alt mode="decision"
+        G->>R: Proposal p1
+        A->>R: Evaluation (each, in order)
+        A->>R: Vote (in order, until settled)
+    else mode="quorum"
+        G->>R: ApprovalRequest r1 (required_approvals = threshold)
+        A->>R: Approve / Reject / Abstain (in order, until settled)
+    end
+    G->>R: Commitment (<action>.approved | <action>.rejected)
+    G-->>G: state[output_key] = ConsensusResult
+```
+
+- **The initiator** is the client's identity. It opens the session, sends the proposal and commits.
+  It never votes and is not counted.
+  - In Decision mode it is listed as the first session participant, because the runtime accepts
+    a `Proposal` only from a declared participant.
+  - In Quorum mode it is not a session participant.
+- **Participants** each send under their own auth. The default is
+  `AuthConfig.for_dev_agent(id)`; pass `auth=AuthConfig.for_bearer(...)` in production.
+- **Evaluation**
+  - An evaluator is an SDK `EvaluationStrategy` (`evaluate(proposal, session_info)`), or a plain
+    function with that signature.
+  - It returns an `EvaluationResult`, or a mapping with `recommendation`, `confidence` and `reason`.
+  - `recommendation` must be one of `APPROVE`, `REVIEW`, `BLOCK` or `REJECT` (case-insensitive).
+    `ABSTAIN` is read as `REVIEW`. Anything else raises.
+  - `proposal` is
+    `{"proposal_id", "option", "rationale", "details", "mode"}`.
+- **Ballots**
+  - `APPROVE` → approve; `REVIEW` → abstain; `REJECT` or `BLOCK` → reject.
+  - In Decision mode a participant may set `voter=`, any SDK `VotingStrategy`, to vote differently
+    from its own recommendation.
+- **Settlement**
+  - Approved once `threshold` participants approve. The default threshold is a strict majority,
+    `n // 2 + 1`.
+  - Rejected as soon as approval is out of reach.
+  - The initiator commits at once, so the remaining participants never vote.
+- **Errors**
+  - If anything fails after the session opens, the node cancels the session and re-raises.
+    This covers a NACK, a transport error, an evaluator exception, and `POLICY_DENIED` on commit.
+  - Nothing is written to state.
+- **Sync and async.** The node is a `RunnableLambda`, so it works under `invoke` and `ainvoke`.
+  The async path runs the blocking gRPC calls in a worker thread.
+
+## Options
+
+| Option | Default | |
+|---|---|---|
+| `client` | required | `MacpClient` authenticated as the initiator |
+| `participants` | required | `ConsensusParticipant(id, evaluator, auth=None, voter=None)`, in voting order |
+| `mode` | `"decision"` | `"decision"` or `"quorum"` |
+| `policy` | `"policy.default"` | a registered `policy_version` (RFC-MACP-0012) |
+| `threshold` | `n // 2 + 1` | approvals required, `1..n` |
+| `proposal` | `state["proposal"]` | `state -> str \| {"option", "rationale"?, "details"?}` |
+| `output_key` | `"macp"` | state key written |
+| `ttl_ms` | `60000` | session TTL |
+| `action` | `"macp.consensus"` | committed as `<action>.approved` / `<action>.rejected` |
+| `initiator` | from the client's auth | the initiator's agent id, needed with a bearer token that has no `sender_hint` |
+
+## What lands in state
+
+```python
+{
+    "outcome": "approve",                       # or "reject"
+    "session_id": "5a5b7088-…",
+    "commitment_hash": "sha256:65f0e9bc…",     # RFC-MACP-0013 macp-commitment-hash/1
+    "mode": "macp.mode.decision.v1",
+    "commitment": {                             # RFC-MACP-0013 §3 projection, plain JSON
+        "commitment_id": "…", "action": "macp.consensus.approved",
+        "authority_scope": "macp.consensus", "reason": "2/3 approvals (threshold 2)",
+        "mode_version": "1.0.0", "policy_version": "policy.default",
+        "configuration_version": "config.default", "outcome_positive": True,
+    },
+    "approvals": 2,
+    "threshold": 2,
+}
+```
+
+`commitment` is plain data, so LangGraph checkpointers can store it.
+
+## Decision receipts
+
+`consensus_receipt(result, issuer=..., issued_at=None, signer=None)` returns an RFC-MACP-0013
+Appendix A `macp-receipt/1` receipt. `issued_at` defaults to now, in `YYYY-MM-DDTHH:MM:SSZ` form.
+
+- With no signer, `verify_receipt(receipt, {})` gives `integrity_only`.
+- Pass `macp_sdk.ReceiptSigner(kid, private_key=<32-byte seed>)` for an Ed25519 attestation.
+  It then verifies `attested` against your trusted keys.
+
+## LLM evaluators
+
+`llm_evaluator(chat_model, system_prompt=None)` turns any LangChain chat model into an evaluator.
+
+- It asks the model for `{"recommendation", "confidence", "reason"}` as JSON. A Markdown code
+  fence around the answer is accepted.
+- An answer it cannot read becomes `REVIEW` with confidence `0` and reason
+  `"unparseable model output"`. That is an abstention, so a confused model cannot approve
+  anything.
+- Tests and the default example use LangChain's `FakeListChatModel` with canned answers, so they
+  need no API key.
+
+## Examples
+
+- [`examples/fraud_review_graph.py`](examples/fraud_review_graph.py)
+  - A fraud-review graph, shaped like macp-playground's LangGraph worker.
+  - It scores signals, then three specialists (fraud, growth, compliance) settle the transaction
+    through the node, and the graph routes on the outcome.
+  - Its LLMs are stubs, so it needs no key and gives the same result every run. The integration
+    test runs it.
+- [`examples/real_llm_review.py`](examples/real_llm_review.py)
+  - The same graph with `claude-opus-5-5` specialists.
+  - Needs `pip install "langgraph-macp[examples]"` (adds `langchain-anthropic`) and
+    `ANTHROPIC_API_KEY`.
+
+## Development
+
+```bash
+cd integrations/langgraph
+python -m venv .venv && .venv/bin/pip install -e "../..[receipts]" -e ".[dev]"
+.venv/bin/pytest tests/unit --cov            # no runtime needed
+MACP_RUNTIME_TARGET=127.0.0.1:50051 .venv/bin/pytest tests/integration
+```
+
+The package lives in the SDK repo but has its own version and release tag, `langgraph-macp-vX.Y.Z`.
+It pins `macp-sdk-python[receipts]>=0.19,<0.20`.
+
+## Other frameworks
+
+- **Claude Agent SDK**: subagents voting through Decision mode. Next in this repo.
+- **OpenAI Agents SDK**: handoffs mapped to MACP Handoff mode. Deferred, planned after the Claude
+  Agent SDK adapter.
+
+## License
+
+Apache-2.0

langgraph_macp-0.1.0.dist-info/RECORD

@@ -0,0 +1,9 @@
+langgraph_macp/__init__.py,sha256=inAtLjk0ZxkMAUYvyJKPkFyMWQIMHrR5vEEySQFR99s,936
+langgraph_macp/llm.py,sha256=N66XeHBjbDxzNFyLxg75saUJX9mSIjutE42YQmmug8U,3342
+langgraph_macp/node.py,sha256=Pbw_g9VMG4eKMEQvhG2n5HH-iHOkdzw7nqjjQKVVQ-k,16928
+langgraph_macp/py.typed,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+langgraph_macp-0.1.0.dist-info/licenses/LICENSE,sha256=xx0jnfkXJvxRnG63LTGOxlggYnIysveWIZ6H3PNdCrQ,11357
+langgraph_macp-0.1.0.dist-info/METADATA,sha256=QYlVfC_Cw9Fs_JP0WYymq3NJXr6gVpauY03-hIFSoMM,10612
+langgraph_macp-0.1.0.dist-info/WHEEL,sha256=mOdXR9C3wMVov7cSY8PY-XCzo3SmmWCHvGQr5csJoao,91
+langgraph_macp-0.1.0.dist-info/top_level.txt,sha256=YtiQnCCbeMpWusSUgrn73VVflzRlACegyk5M24d73Ns,15
+langgraph_macp-0.1.0.dist-info/RECORD,,

langgraph_macp-0.1.0.dist-info/WHEEL

@@ -0,0 +1,5 @@
+Wheel-Version: 1.0
+Generator: setuptools (84.2.1)
+Root-Is-Purelib: true
+Tag: py3-none-any
+

langgraph_macp-0.1.0.dist-info/licenses/LICENSE

@@ -0,0 +1,201 @@
+                                 Apache License
+                           Version 2.0, January 2004
+                        http://www.apache.org/licenses/
+
+   TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+   1. Definitions.
+
+      "License" shall mean the terms and conditions for use, reproduction,
+      and distribution as defined by Sections 1 through 9 of this document.
+
+      "Licensor" shall mean the copyright owner or entity authorized by
+      the copyright owner that is granting the License.
+
+      "Legal Entity" shall mean the union of the acting entity and all
+      other entities that control, are controlled by, or are under common
+      control with that entity. For the purposes of this definition,
+      "control" means (i) the power, direct or indirect, to cause the
+      direction or management of such entity, whether by contract or
+      otherwise, or (ii) ownership of fifty percent (50%) or more of the
+      outstanding shares, or (iii) beneficial ownership of such entity.
+
+      "You" (or "Your") shall mean an individual or Legal Entity
+      exercising permissions granted by this License.
+
+      "Source" form shall mean the preferred form for making modifications,
+      including but not limited to software source code, documentation
+      source, and configuration files.
+
+      "Object" form shall mean any form resulting from mechanical
+      transformation or translation of a Source form, including but
+      not limited to compiled object code, generated documentation,
+      and conversions to other media types.
+
+      "Work" shall mean the work of authorship, whether in Source or
+      Object form, made available under the License, as indicated by a
+      copyright notice that is included in or attached to the work
+      (an example is provided in the Appendix below).
+
+      "Derivative Works" shall mean any work, whether in Source or Object
+      form, that is based on (or derived from) the Work and for which the
+      editorial revisions, annotations, elaborations, or other modifications
+      represent, as a whole, an original work of authorship. For the purposes
+      of this License, Derivative Works shall not include works that remain
+      separable from, or merely link (or bind by name) to the interfaces of,
+      the Work and Derivative Works thereof.
+
+      "Contribution" shall mean any work of authorship, including
+      the original version of the Work and any modifications or additions
+      to that Work or Derivative Works thereof, that is intentionally
+      submitted to Licensor for inclusion in the Work by the copyright owner
+      or by an individual or Legal Entity authorized to submit on behalf of
+      the copyright owner. For the purposes of this definition, "submitted"
+      means any form of electronic, verbal, or written communication sent
+      to the Licensor or its representatives, including but not limited to
+      communication on electronic mailing lists, source code control systems,
+      and issue tracking systems that are managed by, or on behalf of, the
+      Licensor for the purpose of discussing and improving the Work, but
+      excluding communication that is conspicuously marked or otherwise
+      designated in writing by the copyright owner as "Not a Contribution."
+
+      "Contributor" shall mean Licensor and any individual or Legal Entity
+      on behalf of whom a Contribution has been received by Licensor and
+      subsequently incorporated within the Work.
+
+   2. Grant of Copyright License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      copyright license to reproduce, prepare Derivative Works of,
+      publicly display, publicly perform, sublicense, and distribute the
+      Work and such Derivative Works in Source or Object form.
+
+   3. Grant of Patent License. Subject to the terms and conditions of
+      this License, each Contributor hereby grants to You a perpetual,
+      worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+      (except as stated in this section) patent license to make, have made,
+      use, offer to sell, sell, import, and otherwise transfer the Work,
+      where such license applies only to those patent claims licensable
+      by such Contributor that are necessarily infringed by their
+      Contribution(s) alone or by combination of their Contribution(s)
+      with the Work to which such Contribution(s) was submitted. If You
+      institute patent litigation against any entity (including a
+      cross-claim or counterclaim in a lawsuit) alleging that the Work
+      or a Contribution incorporated within the Work constitutes direct
+      or contributory patent infringement, then any patent licenses
+      granted to You under this License for that Work shall terminate
+      as of the date such litigation is filed.
+
+   4. Redistribution. You may reproduce and distribute copies of the
+      Work or Derivative Works thereof in any medium, with or without
+      modifications, and in Source or Object form, provided that You
+      meet the following conditions:
+
+      (a) You must give any other recipients of the Work or
+          Derivative Works a copy of this License; and
+
+      (b) You must cause any modified files to carry prominent notices
+          stating that You changed the files; and
+
+      (c) You must retain, in the Source form of any Derivative Works
+          that You distribute, all copyright, patent, trademark, and
+          attribution notices from the Source form of the Work,
+          excluding those notices that do not pertain to any part of
+          the Derivative Works; and
+
+      (d) If the Work includes a "NOTICE" text file as part of its
+          distribution, then any Derivative Works that You distribute must
+          include a readable copy of the attribution notices contained
+          within such NOTICE file, excluding those notices that do not
+          pertain to any part of the Derivative Works, in at least one
+          of the following places: within a NOTICE text file distributed
+          as part of the Derivative Works; within the Source form or
+          documentation, if provided along with the Derivative Works; or,
+          within a display generated by the Derivative Works, if and
+          wherever such third-party notices normally appear. The contents
+          of the NOTICE file are for informational purposes only and
+          do not modify the License. You may add Your own attribution
+          notices within Derivative Works that You distribute, alongside
+          or as an addendum to the NOTICE text from the Work, provided
+          that such additional attribution notices cannot be construed
+          as modifying the License.
+
+      You may add Your own copyright statement to Your modifications and
+      may provide additional or different license terms and conditions
+      for use, reproduction, or distribution of Your modifications, or
+      for any such Derivative Works as a whole, provided Your use,
+      reproduction, and distribution of the Work otherwise complies with
+      the conditions stated in this License.
+
+   5. Submission of Contributions. Unless You explicitly state otherwise,
+      any Contribution intentionally submitted for inclusion in the Work
+      by You to the Licensor shall be under the terms and conditions of
+      this License, without any additional terms or conditions.
+      Notwithstanding the above, nothing herein shall supersede or modify
+      the terms of any separate license agreement you may have executed
+      with Licensor regarding such Contributions.
+
+   6. Trademarks. This License does not grant permission to use the trade
+      names, trademarks, service marks, or product names of the Licensor,
+      except as required for reasonable and customary use in describing the
+      origin of the Work and reproducing the content of the NOTICE file.
+
+   7. Disclaimer of Warranty. Unless required by applicable law or
+      agreed to in writing, Licensor provides the Work (and each
+      Contributor provides its Contributions) on an "AS IS" BASIS,
+      WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+      implied, including, without limitation, any warranties or conditions
+      of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+      PARTICULAR PURPOSE. You are solely responsible for determining the
+      appropriateness of using or redistributing the Work and assume any
+      risks associated with Your exercise of permissions under this License.
+
+   8. Limitation of Liability. In no event and under no legal theory,
+      whether in tort (including negligence), contract, or otherwise,
+      unless required by applicable law (such as deliberate and grossly
+      negligent acts) or agreed to in writing, shall any Contributor be
+      liable to You for damages, including any direct, indirect, special,
+      incidental, or consequential damages of any character arising as a
+      result of this License or out of the use or inability to use the
+      Work (including but not limited to damages for loss of goodwill,
+      work stoppage, computer failure or malfunction, or any and all
+      other commercial damages or losses), even if such Contributor
+      has been advised of the possibility of such damages.
+
+   9. Accepting Warranty or Additional Liability. While redistributing
+      the Work or Derivative Works thereof, You may choose to offer,
+      and charge a fee for, acceptance of support, warranty, indemnity,
+      or other liability obligations and/or rights consistent with this
+      License. However, in accepting such obligations, You may act only
+      on Your own behalf and on Your sole responsibility, not on behalf
+      of any other Contributor, and only if You agree to indemnify,
+      defend, and hold each Contributor harmless for any liability
+      incurred by, or claims asserted against, such Contributor by reason
+      of your accepting any such warranty or additional liability.
+
+   END OF TERMS AND CONDITIONS
+
+   APPENDIX: How to apply the Apache License to your work.
+
+      To apply the Apache License to your work, attach the following
+      boilerplate notice, with the fields enclosed by brackets "[]"
+      replaced with your own identifying information. (Don't include
+      the brackets!)  The text should be enclosed in the appropriate
+      comment syntax for the file format. We also recommend that a
+      file or class name and description of purpose be included on the
+      same "printed page" as the copyright notice for easier
+      identification within third-party archives.
+
+   Copyright [yyyy] [name of copyright owner]
+
+   Licensed under the Apache License, Version 2.0 (the "License");
+   you may not use this file except in compliance with the License.
+   You may obtain a copy of the License at
+
+       http://www.apache.org/licenses/LICENSE-2.0
+
+   Unless required by applicable law or agreed to in writing, software
+   distributed under the License is distributed on an "AS IS" BASIS,
+   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+   See the License for the specific language governing permissions and
+   limitations under the License.

langgraph_macp-0.1.0.dist-info/top_level.txt

@@ -0,0 +1 @@
+langgraph_macp