mirror of
https://github.com/razzant/ouroboros.git
synced 2026-10-19 10:04:16 +00:00
4a4fbb73b3
Make headless Google Colab a first-class Ouroboros runtime controlled entirely over Telegram, and treat reviewed transport/control skills as full replacements for the local UI. - Colab source-mode bootstrap: mounts Drive for persistent data/, role-based remotes (managed=official, origin=personal fork), merges existing Drive settings on re-run, and installs+enables the Telegram bridge in full_access over the loopback gateway. - Role-based GitHub remotes with auto-fork personal origin while preserving the official managed update remote; a clone-default official origin is reclassified as managed (never an origin conflict) and configure_personal_remote guarantees managed exists before repointing origin. - Transport/control-skill doctrine: a reviewed + token-bound + enabled + owner-granted + owner-bound chat/control skill is a first-class control surface; review judges actual safety properties, not breadth of control (CHECKLISTS, CREATING_SKILLS, ARCHITECTURE). - Owner-external slash TOFU so a real Telegram owner is authorized without ever locking out the local web owner; external /review targets the requesting chat. - Adaptive LLM request-parameter normalization (OpenRouter no longer drops Opus on unsupported sampling params). - official_hub review profile downgrades severity-driven hygiene/bug findings to warnings for hash-verified official OuroborosHub payloads (re-verified via the local sidecar on reload); trust-boundary items and deterministic gates stay fail-closed. Deterministic preflight failures persist+aggregate to PENDING (non-executable under every enforcement mode, surviving reload). - Reviewed-skill auto-grants default on; rescue snapshots recoverable via real git refs; ff-only restore of personal-origin self-mods across Colab sessions.
338 lines
15 KiB
Python
338 lines
15 KiB
Python
"""Tests for effort, review models, and review enforcement settings."""
|
|
import json
|
|
import os
|
|
from ouroboros.config import (
|
|
SETTINGS_DEFAULTS,
|
|
apply_settings_to_env,
|
|
resolve_effort,
|
|
get_review_models,
|
|
get_review_enforcement,
|
|
get_scope_review_models,
|
|
get_task_review_mode,
|
|
get_auto_grant_enabled,
|
|
)
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# Legacy env var backward compat
|
|
# ---------------------------------------------------------------------------
|
|
|
|
def test_initial_effort_default(monkeypatch):
|
|
"""Default effort is 'medium' when env var not set."""
|
|
monkeypatch.delenv("OUROBOROS_EFFORT_TASK", raising=False)
|
|
assert resolve_effort("task") == "medium"
|
|
|
|
|
|
def test_initial_effort_valid_values(monkeypatch):
|
|
"""Valid effort values pass through unchanged via OUROBOROS_EFFORT_TASK."""
|
|
for effort in ("none", "low", "medium", "high"):
|
|
monkeypatch.setenv("OUROBOROS_EFFORT_TASK", effort)
|
|
assert resolve_effort("task") == effort
|
|
|
|
|
|
def test_initial_effort_invalid_falls_back_to_medium(monkeypatch):
|
|
"""Invalid effort values fall back to 'medium'."""
|
|
monkeypatch.setenv("OUROBOROS_EFFORT_TASK", "extreme")
|
|
assert resolve_effort("task") == "medium"
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# New per-type defaults in SETTINGS_DEFAULTS
|
|
# ---------------------------------------------------------------------------
|
|
|
|
def test_effort_defaults_in_config():
|
|
"""All effort keys have correct defaults in SETTINGS_DEFAULTS."""
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_EFFORT_TASK") == "medium"
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_EFFORT_EVOLUTION") == "high"
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_EFFORT_REVIEW") == "medium"
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_EFFORT_SCOPE_REVIEW") == "high"
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_EFFORT_CONSCIOUSNESS") == "high"
|
|
|
|
|
|
def test_review_models_default_in_config():
|
|
"""OUROBOROS_REVIEW_MODELS has a default value in config."""
|
|
val = SETTINGS_DEFAULTS.get("OUROBOROS_REVIEW_MODELS", "")
|
|
assert val # non-empty
|
|
models = [m.strip() for m in val.split(",") if m.strip()]
|
|
assert len(models) >= 2 # quorum requires at least 2
|
|
|
|
|
|
def test_review_enforcement_default_in_config():
|
|
"""OUROBOROS_REVIEW_ENFORCEMENT defaults to advisory."""
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_REVIEW_ENFORCEMENT") == "advisory"
|
|
|
|
|
|
def test_scope_review_and_task_review_defaults_in_config():
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_SCOPE_REVIEW_MODELS") == "openai/gpt-5.5"
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_TASK_REVIEW_MODE") == "auto"
|
|
|
|
|
|
def test_auto_grant_reviewed_skills_default_in_config():
|
|
assert SETTINGS_DEFAULTS.get("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS") == "true"
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# get_review_models() — single source of truth
|
|
# ---------------------------------------------------------------------------
|
|
|
|
def test_get_review_models_default(monkeypatch):
|
|
"""get_review_models() returns the config default when env is unset."""
|
|
monkeypatch.delenv("OUROBOROS_REVIEW_MODELS", raising=False)
|
|
monkeypatch.delenv("OPENAI_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
|
|
monkeypatch.delenv("OUROBOROS_MODEL", raising=False)
|
|
models = get_review_models()
|
|
assert isinstance(models, list)
|
|
assert len(models) >= 2
|
|
assert all("/" in m for m in models) # valid OpenRouter model IDs
|
|
|
|
|
|
def test_get_review_models_custom(monkeypatch):
|
|
"""get_review_models() returns custom models when env is set."""
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_MODELS", "a/b,c/d")
|
|
monkeypatch.delenv("OPENAI_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
|
|
monkeypatch.delenv("OUROBOROS_MODEL", raising=False)
|
|
models = get_review_models()
|
|
assert models == ["a/b", "c/d"]
|
|
|
|
|
|
def test_get_review_models_empty_env_falls_back_to_default(monkeypatch):
|
|
"""get_review_models() falls back to default when env is empty string."""
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_MODELS", "")
|
|
monkeypatch.delenv("OPENAI_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
|
|
monkeypatch.delenv("OUROBOROS_MODEL", raising=False)
|
|
models = get_review_models()
|
|
# Must return the default, not an empty list
|
|
assert len(models) >= 2
|
|
assert models == [m.strip() for m in SETTINGS_DEFAULTS["OUROBOROS_REVIEW_MODELS"].split(",") if m.strip()]
|
|
|
|
|
|
def test_get_review_models_falls_back_to_main_light_light_in_openai_only_mode(monkeypatch):
|
|
"""v4.39.0: direct-provider fallback returns [main, light, light] (3 slots,
|
|
2 unique) instead of the legacy [main]*N so both commit triad and
|
|
plan_task have a quorum-safe reviewer list out of the box. The light slot
|
|
picks up the provider default (OPENAI_DIRECT_DEFAULTS['light'] =
|
|
openai::gpt-5.5-mini) when OUROBOROS_MODEL_LIGHT is not explicitly set."""
|
|
monkeypatch.setenv("OPENAI_API_KEY", "sk-openai")
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.delenv("OUROBOROS_MODEL_LIGHT", raising=False)
|
|
monkeypatch.setenv("OUROBOROS_MODEL", "openai::gpt-5.5")
|
|
monkeypatch.setenv(
|
|
"OUROBOROS_REVIEW_MODELS",
|
|
"openai/gpt-5.5,google/gemini-3.5-flash,anthropic/claude-opus-4.6",
|
|
)
|
|
|
|
models = get_review_models()
|
|
|
|
assert models == [
|
|
"openai::gpt-5.5",
|
|
"openai::gpt-5.5-mini",
|
|
"openai::gpt-5.5-mini",
|
|
]
|
|
|
|
|
|
def test_get_review_models_preserves_explicit_official_openai_list(monkeypatch):
|
|
monkeypatch.setenv("OPENAI_API_KEY", "sk-openai")
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("ANTHROPIC_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.setenv("OUROBOROS_MODEL", "openai::gpt-5.5")
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_MODELS", "openai/gpt-5.5,openai/gpt-4.1")
|
|
|
|
models = get_review_models()
|
|
|
|
assert models == ["openai::gpt-5.5", "openai::gpt-4.1"]
|
|
|
|
|
|
def test_get_review_models_falls_back_to_main_light_light_in_anthropic_only_mode(monkeypatch):
|
|
"""v4.39.0: same direct-provider fallback as OpenAI — [main, light, light]
|
|
with light = ANTHROPIC_DIRECT_DEFAULTS['light'] = anthropic::claude-sonnet-4-6
|
|
when OUROBOROS_MODEL_LIGHT is not explicitly set."""
|
|
monkeypatch.setenv("ANTHROPIC_API_KEY", "sk-ant")
|
|
monkeypatch.delenv("OPENROUTER_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_API_KEY", raising=False)
|
|
monkeypatch.delenv("OPENAI_BASE_URL", raising=False)
|
|
monkeypatch.delenv("OPENAI_COMPATIBLE_API_KEY", raising=False)
|
|
monkeypatch.delenv("CLOUDRU_FOUNDATION_MODELS_API_KEY", raising=False)
|
|
monkeypatch.delenv("OUROBOROS_MODEL_LIGHT", raising=False)
|
|
monkeypatch.setenv("OUROBOROS_MODEL", "anthropic::claude-opus-4-6")
|
|
monkeypatch.setenv(
|
|
"OUROBOROS_REVIEW_MODELS",
|
|
"openai/gpt-5.5,google/gemini-3.5-flash,anthropic/claude-opus-4.6",
|
|
)
|
|
|
|
models = get_review_models()
|
|
|
|
assert models == [
|
|
"anthropic::claude-opus-4-6",
|
|
"anthropic::claude-sonnet-4-6",
|
|
"anthropic::claude-sonnet-4-6",
|
|
]
|
|
|
|
|
|
def test_get_review_enforcement_default(monkeypatch):
|
|
"""get_review_enforcement() returns the config default when env is unset."""
|
|
monkeypatch.delenv("OUROBOROS_REVIEW_ENFORCEMENT", raising=False)
|
|
assert get_review_enforcement() == "advisory"
|
|
|
|
|
|
def test_get_review_enforcement_custom(monkeypatch):
|
|
"""get_review_enforcement() accepts advisory and blocking."""
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_ENFORCEMENT", "advisory")
|
|
assert get_review_enforcement() == "advisory"
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_ENFORCEMENT", "blocking")
|
|
assert get_review_enforcement() == "blocking"
|
|
|
|
|
|
def test_get_review_enforcement_invalid_falls_back(monkeypatch):
|
|
"""Unknown values fall back to advisory (the default)."""
|
|
monkeypatch.setenv("OUROBOROS_REVIEW_ENFORCEMENT", "strictest")
|
|
assert get_review_enforcement() == "advisory"
|
|
|
|
|
|
def test_get_scope_review_models_preserves_duplicate_slots(monkeypatch):
|
|
monkeypatch.setenv("OUROBOROS_SCOPE_REVIEW_MODELS", "model/a, model/a, model/b")
|
|
assert get_scope_review_models() == ["model/a", "model/a", "model/b"]
|
|
|
|
|
|
def test_get_scope_review_models_falls_back_to_singular(monkeypatch):
|
|
monkeypatch.setenv("OUROBOROS_SCOPE_REVIEW_MODELS", "")
|
|
monkeypatch.setenv("OUROBOROS_SCOPE_REVIEW_MODEL", "legacy/scope")
|
|
assert get_scope_review_models() == ["legacy/scope"]
|
|
|
|
|
|
def test_get_task_review_mode_clamps_invalid(monkeypatch):
|
|
monkeypatch.delenv("OUROBOROS_TASK_REVIEW_MODE", raising=False)
|
|
assert get_task_review_mode() == "auto"
|
|
monkeypatch.setenv("OUROBOROS_TASK_REVIEW_MODE", "required")
|
|
assert get_task_review_mode() == "required"
|
|
monkeypatch.setenv("OUROBOROS_TASK_REVIEW_MODE", "blocking")
|
|
assert get_task_review_mode() == "auto"
|
|
|
|
|
|
def test_get_auto_grant_enabled(monkeypatch, tmp_path):
|
|
from ouroboros import config as cfg
|
|
|
|
monkeypatch.setattr(cfg, "SETTINGS_PATH", tmp_path / "missing-settings.json", raising=True)
|
|
monkeypatch.delenv("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS", raising=False)
|
|
# New SSOT default is "true": absent settings file + absent env → enabled.
|
|
assert cfg.get_auto_grant_enabled() is True
|
|
monkeypatch.setenv("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS", "true")
|
|
assert cfg.get_auto_grant_enabled() is True
|
|
monkeypatch.setenv("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS", "false")
|
|
assert cfg.get_auto_grant_enabled() is False
|
|
|
|
|
|
def test_get_auto_grant_enabled_prefers_settings_file(monkeypatch, tmp_path):
|
|
from ouroboros import config as cfg
|
|
|
|
settings_path = tmp_path / "settings.json"
|
|
settings_path.write_text(
|
|
json.dumps({"OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS": "true"}),
|
|
encoding="utf-8",
|
|
)
|
|
monkeypatch.setattr(cfg, "SETTINGS_PATH", settings_path, raising=True)
|
|
monkeypatch.setenv("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS", "false")
|
|
|
|
assert cfg.get_auto_grant_enabled() is True
|
|
|
|
|
|
def test_apply_settings_clears_review_models_restores_default(monkeypatch):
|
|
"""Clearing OUROBOROS_REVIEW_MODELS in settings restores the default in env."""
|
|
# Simulate user clearing the field in Settings UI (empty string)
|
|
settings = {"OUROBOROS_REVIEW_MODELS": ""}
|
|
apply_settings_to_env(settings)
|
|
# env var should be the default, not empty
|
|
env_val = os.environ.get("OUROBOROS_REVIEW_MODELS", "")
|
|
assert env_val == SETTINGS_DEFAULTS["OUROBOROS_REVIEW_MODELS"]
|
|
# get_review_models() should also return correct defaults
|
|
assert len(get_review_models()) >= 2
|
|
|
|
|
|
def test_apply_settings_clears_review_enforcement_restores_default(monkeypatch):
|
|
"""Clearing OUROBOROS_REVIEW_ENFORCEMENT restores the default in env."""
|
|
settings = {"OUROBOROS_REVIEW_ENFORCEMENT": ""}
|
|
apply_settings_to_env(settings)
|
|
env_val = os.environ.get("OUROBOROS_REVIEW_ENFORCEMENT", "")
|
|
assert env_val == SETTINGS_DEFAULTS["OUROBOROS_REVIEW_ENFORCEMENT"]
|
|
assert get_review_enforcement() == "advisory"
|
|
|
|
|
|
def test_apply_settings_clears_task_and_scope_review_restores_default(monkeypatch):
|
|
settings = {"OUROBOROS_SCOPE_REVIEW_MODELS": "", "OUROBOROS_SCOPE_REVIEW_MODEL": "", "OUROBOROS_TASK_REVIEW_MODE": ""}
|
|
apply_settings_to_env(settings)
|
|
assert os.environ.get("OUROBOROS_SCOPE_REVIEW_MODELS") == SETTINGS_DEFAULTS["OUROBOROS_SCOPE_REVIEW_MODELS"]
|
|
assert os.environ.get("OUROBOROS_TASK_REVIEW_MODE") == SETTINGS_DEFAULTS["OUROBOROS_TASK_REVIEW_MODE"]
|
|
|
|
|
|
# ---------------------------------------------------------------------------
|
|
# apply_settings_to_env propagation
|
|
# ---------------------------------------------------------------------------
|
|
|
|
def test_apply_settings_to_env_includes_effort_keys(monkeypatch, tmp_path):
|
|
"""apply_settings_to_env propagates all effort keys."""
|
|
settings = {
|
|
"OUROBOROS_EFFORT_TASK": "low",
|
|
"OUROBOROS_EFFORT_EVOLUTION": "medium",
|
|
"OUROBOROS_EFFORT_REVIEW": "high",
|
|
"OUROBOROS_EFFORT_SCOPE_REVIEW": "low",
|
|
"OUROBOROS_EFFORT_CONSCIOUSNESS": "none",
|
|
"OUROBOROS_REVIEW_MODELS": "model-a,model-b",
|
|
"OUROBOROS_REVIEW_ENFORCEMENT": "advisory",
|
|
"OUROBOROS_SCOPE_REVIEW_MODELS": "scope-a,scope-b",
|
|
"OUROBOROS_TASK_REVIEW_MODE": "required",
|
|
"OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS": "true",
|
|
"OUROBOROS_RETURN_REASONING": "",
|
|
}
|
|
apply_settings_to_env(settings)
|
|
assert os.environ.get("OUROBOROS_EFFORT_TASK") == "low"
|
|
assert os.environ.get("OUROBOROS_EFFORT_EVOLUTION") == "medium"
|
|
assert os.environ.get("OUROBOROS_EFFORT_REVIEW") == "high"
|
|
assert os.environ.get("OUROBOROS_EFFORT_SCOPE_REVIEW") == "low"
|
|
assert os.environ.get("OUROBOROS_EFFORT_CONSCIOUSNESS") == "none"
|
|
assert os.environ.get("OUROBOROS_REVIEW_MODELS") == "model-a,model-b"
|
|
assert os.environ.get("OUROBOROS_REVIEW_ENFORCEMENT") == "advisory"
|
|
assert os.environ.get("OUROBOROS_SCOPE_REVIEW_MODELS") == "scope-a,scope-b"
|
|
assert os.environ.get("OUROBOROS_TASK_REVIEW_MODE") == "required"
|
|
assert os.environ.get("OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS") == "true"
|
|
assert os.environ.get("OUROBOROS_RETURN_REASONING") == ""
|
|
# cleanup
|
|
for k in ("OUROBOROS_EFFORT_TASK", "OUROBOROS_EFFORT_EVOLUTION",
|
|
"OUROBOROS_EFFORT_REVIEW", "OUROBOROS_EFFORT_SCOPE_REVIEW",
|
|
"OUROBOROS_EFFORT_CONSCIOUSNESS",
|
|
"OUROBOROS_REVIEW_MODELS", "OUROBOROS_REVIEW_ENFORCEMENT",
|
|
"OUROBOROS_SCOPE_REVIEW_MODELS", "OUROBOROS_TASK_REVIEW_MODE",
|
|
"OUROBOROS_AUTO_GRANT_REVIEWED_SKILLS", "OUROBOROS_RETURN_REASONING"):
|
|
os.environ.pop(k, None)
|
|
|
|
import ouroboros.config as cfg
|
|
|
|
settings_path = tmp_path / "settings.json"
|
|
settings_path.write_text(json.dumps({"OUROBOROS_RETURN_REASONING": True}), encoding="utf-8")
|
|
monkeypatch.setattr(cfg, "SETTINGS_PATH", settings_path, raising=True)
|
|
monkeypatch.setenv("OUROBOROS_RETURN_REASONING", "")
|
|
|
|
loaded = cfg.load_settings()
|
|
assert loaded["OUROBOROS_RETURN_REASONING"] == ""
|
|
cfg.apply_settings_to_env(loaded)
|
|
assert os.environ.get("OUROBOROS_RETURN_REASONING") == ""
|