## Summary Adds `experimental/`, a first version of worker RSI: a Curator that shapes a Raven worker's harness from the task, the materials handed to it and feedback, the way a new employee is trained. The worker is a harness-of-harnesses. A root Raven harness faces the user and calls child harnesses, each a full Raven agent hosted over ACP, through the nodes of its playbooks. Every harness has the same four strategy facets (memory, planning, capability, action). The Curator reads the actual harnesses and generates code and config on those facets at both levels: the root candidate sets the playbooks and writes per-node requirements, each child is curated against the requirements of its nodes, and the set is checked on copies and activated together. Layout: - `curator/`: staged generation (understand, select, design, implement, repair), two-level composition, harness contracts, and `raven_adapter/`, the only layer that touches a worker's Raven runtime - `analyst/`: turns an evaluator's words and execution records into observable behavior requirements - `iteration/`: the generic loop (onboarding curation, then trial, review and curation each round) and a CLI for a human evaluator - `simulation/`: a simulated travel agency that runs the loop, plus one showcase case (`simulation/cases/s0925c/`: case notes, the exported transcript split by phase, and the round 1 and round 3 decks) - `docs/design.md` (structure, curation flow, composition sequence, the Raven connection) and `docs/rsi-iteration.md` (the loop and the Analyst) Key decisions: - No change under `raven/`, and `raven/` does not import `experimental/`. Generated artifacts land on Raven's existing extension points. The few Raven internals the adapter reads are listed in `docs/design.md` section 5.4. - The Curator explores a read-only source snapshot; the caller names the evaluation side to withhold from it, so the generic layers name no scenario. - The single-layer owner analysis is kept only to reproduce the showcase case and is marked for removal; the two-layer Analyst is the default. - The two decks are `.pptx` files of about 230 KB each. They pass `check-large-files`; flagging them in case they count as report assets under AGENTS.md section 7, in which case they can move out of git behind a link. - One follow-up for Raven 0.2.2: playbooks in prompt and rounds modes ship no fixed nodes, so node requirements are read from dag playbooks only. ## Type - [ ] Fix - [x] Feature - [ ] Docs - [ ] CI / tooling - [ ] Refactor - [ ] Other ## Verification On the branch rebased onto the latest `main`: - `uv run pytest tests/test_harness_curator_*.py tests/test_analyst_*.py tests/test_iteration_*.py tests/test_simulation_*.py -q`: 368 passed - `uv run pytest -m integration tests/integration/test_harness_curator_*.py --deselect tests/integration/test_harness_curator_real_llm.py -q` (offline, stub providers): 57 passed - `uv run pytest -q` (whole suite, with the host proxy variables unset and `PYTHONTZPATH` set): 26692 passed, 2 failed. One was an order-dependent test of this PR and is fixed here; the other, `tests/test_subagent_node_runtime.py::test_what_cannot_be_read_names_nothing`, fails only because the machine runs as root and is unrelated. - `uv run ruff check` and `uv run ruff format --check` on the touched files: clean - `scripts/check_large_files.py origin/main..HEAD`, `scripts/check_source_language.py origin/main..HEAD`, `scripts/check_commit_messages.py origin/main..HEAD` and `npx commitlint --from origin/main --to HEAD`: all exit 0 - [x] Relevant tests pass locally - [x] Relevant lint / type checks pass locally - [x] User-facing docs or screenshots are updated when needed ## Risk No user-visible change for Raven: `experimental/` is not imported by `raven/` and is not packaged in the wheel (`packages = ["raven"]`). `CONTEXT.md` gains one glossary entry. The Curator's exploration offers a shell only under an OS sandbox executor. Rollback is a revert of the squash commit; nothing else depends on `experimental/`. - [x] Security impact considered - [x] Backward compatibility considered - [x] Rollback path is clear for risky changes ## Related Issues N/A --------- Co-authored-by: Yuqi Yang <yiqiyang524@gmail.com> Co-authored-by: Claude (claude-opus-5-5) <noreply@anthropic.com>
78 lines
3.5 KiB
Python
78 lines
3.5 KiB
Python
"""Drill cards with drawn values: every playing meets new facts, yet each card keeps the path it was written for."""
|
|
|
|
import random
|
|
|
|
import pytest
|
|
|
|
from experimental.simulation.cards import draw, split
|
|
from experimental.simulation.reference import Rules
|
|
from experimental.simulation.scenario import BUNDLED, Scenario
|
|
from experimental.simulation.traveller import Traveller
|
|
from raven.contracts.llm_provider import LLMResponse
|
|
|
|
TRAVEL = BUNDLED / "travel_agency"
|
|
DRAWS = 300
|
|
|
|
|
|
class Provider:
|
|
def __init__(self):
|
|
self.requests = []
|
|
|
|
async def chat_with_retry(self, **kwargs):
|
|
self.requests.append(kwargs)
|
|
return LLMResponse(content="Hello")
|
|
|
|
|
|
def chosen(rules: Rules, trip) -> tuple[str, int]:
|
|
"""The product the SOP's selection matrix leads to (S3.1, S3.3): by nightly budget, then down while it does not fit."""
|
|
order = list(rules.products)
|
|
nightly = trip.budget / trip.persons / trip.nights
|
|
tier = 2 if nightly > 2000 else 1 if nightly > 500 else 0
|
|
while tier >= 0:
|
|
product = order[tier]
|
|
expected = rules.expected(product, trip)
|
|
if rules.products[product].base <= trip.nights and expected and expected.total <= trip.budget:
|
|
return product, tier
|
|
tier -= 1
|
|
raise AssertionError(f"no product fits {trip}")
|
|
|
|
|
|
def test_every_card_draws_fresh_facts_but_keeps_its_product_season_and_budget_fit():
|
|
scenario = Scenario.load(TRAVEL)
|
|
rules = Rules.load(scenario)
|
|
tiers = {}
|
|
for persona in scenario.personas:
|
|
played = [persona.play(random.Random(f"7:{persona.name}:{count}")) for count in range(DRAWS)]
|
|
assert all(card.trip and "{" not in card.text and card.trip.phone in card.text for card in played)
|
|
assert all(f"{card.trip.start[0]}" in card.text and str(card.trip.budget) in card.text for card in played)
|
|
assert len({(card.trip.start, card.trip.phone, card.trip.budget) for card in played}) > DRAWS * 0.9
|
|
assert len({rules.season(card.trip.start) for card in played}) == 1
|
|
assert len({card.trip.persons for card in played}) == 1
|
|
tiers[persona.name] = {chosen(rules, card.trip)[1] for card in played}
|
|
assert tiers == {"family": {0}, "premium": {2}, "professional": {1}, "returning": {0}, "student": {0}}
|
|
|
|
|
|
def test_one_seed_draws_the_same_facts_and_a_card_without_values_plays_as_written():
|
|
values, body = split("---\nvalues:\n who: [Ann, Bo]\n n: {from: 1, to: 9}\n---\n{who} has {n}.\n")
|
|
first, again = (draw("x", values, body, random.Random("s")) for _ in range(2))
|
|
assert first == again and first.trip is None and "{" not in first.text
|
|
assert split("Plain card.") == ({}, "Plain card.")
|
|
assert draw("plain", {}, "Keep {braces}.", random.Random(1)).text == "Keep {braces}."
|
|
trip = {"start": {"from": "10-30", "to": "10-30"}, "nights": 3}
|
|
with pytest.raises(ValueError, match="phone"):
|
|
draw("x", trip, "", random.Random(1))
|
|
|
|
|
|
async def test_the_traveller_plays_its_card_afresh_each_conversation_and_one_seed_repeats_the_rounds():
|
|
scenario = Scenario.load(TRAVEL)
|
|
student = next(persona for persona in scenario.personas if persona.name == "student")
|
|
traveller = Traveller(student, Provider(), seed=3)
|
|
await traveller.speak([])
|
|
first = traveller.card
|
|
await traveller.speak([])
|
|
assert traveller.card != first and traveller.played == 2
|
|
again = Traveller(student, Provider(), seed=3)
|
|
await again.speak([])
|
|
assert again.card == first
|
|
assert first.trip.end == (first.trip.start[0], first.trip.start[1] + 3)
|