Hello-QM/catgo-LRG / server/catgo/workflow
catgo-workflow
VerifiedCreate and manage computational chemistry workflows with CatGo. Supports VASP, CP2K, ORCA, MLP, LAMMPS. Build OER/HER/CO2RR workflows, geometry optimization, frequency analysis, Gibbs energy calculations.
config.py
View on GitHub"""Layered configuration system.
Resolution order (highest priority wins):
Task params → Workflow config → User config (~/.catgo/config.yaml) → System defaults
Environment variable override: CATGO_ENGINE_POLL_INTERVAL=10
"""
from __future__ import annotations
import copy
import os
from pathlib import Path
from typing import Any
DEFAULT_CONFIG: dict[str, Any] = {
"engine": {
"poll_interval": 30,
"submit_batch_size": 5,
"max_concurrent_jobs": 20,
"result_collect_timeout": 300,
},
"hpc": {
"ssh_timeout": 30,
"ssh_retry_max": 3,
"ssh_retry_backoff": 10,
"poll_retry_max": 5,
"poll_retry_backoff": 60,
"poll_retry_factor": 2,
},
"retry": {
"max_retries": 3,
"backoff_base": 60,
"backoff_factor": 2,
"max_backoff": 3600,
},
"defaults": {
"vasp": {
"ENCUT": 520, "EDIFF": 1e-5, "PREC": "Accurate", "ALGO": "Fast",
"ISMEAR": 0, "SIGMA": 0.05, "LREAL": "Auto", "NELM": 200,
"ISPIN": 1, "LORBIT": 11, "LWAVE": False, "LCHARG": False, "NCORE": 4,
},
"vasp_geo_opt": {
"ISIF": 2, "NSW": 200, "EDIFFG": -0.02, "IBRION": 2,
},
"vasp_freq": {
"IBRION": 5, "NFREE": 2, "POTIM": 0.015, "LREAL": ".FALSE.", "EDIFF": 1e-6,
},
"vasp_single_point": {
"NSW": 0, "IBRION": -1, "NEDOS": 3001,
},
"cp2k": {
"cutoff": 600, "rel_cutoff": 60, "xc_functional": "PBE", "scf_max_iter": 200,
},
"orca": {
"method": "B3LYP", "basis_set": "def2-SVP",
"charge": 0, "multiplicity": 1, "num_cores": 4, "max_core_mb": 4000,
"max_iterations": 100,
},
"gibbs": {
"temperature": 298.15, "freq_cutoff": 50, "pressure_atm": 1.0, "phase": "adsorbed",
},
},
"paths": {
"work_dir_template": "{base_dir}/{workflow_id}/{task_id}",
"base_dir": "",
"db_path": "~/.catgo/catgo.db",
"log_dir": "~/.catgo/logs/",
"config_dir": "~/.catgo/",
},
"logging": {
"level": "INFO",
"max_log_size": 10485760,
"log_rotation": 5,
},
}
def _deep_merge(base: dict, override: dict) -> dict:
"""Merge override into base recursively. Override wins on conflicts."""
result = copy.deepcopy(base)
for key, value in override.items():
if key in result and isinstance(result[key], dict) and isinstance(value, dict):
result[key] = _deep_merge(result[key], value)
else:
result[key] = copy.deepcopy(value)
return result
def _apply_env_vars(config: dict, prefix: str = "CATGO") -> dict:
"""Override config values from environment variables.
CATGO_ENGINE_POLL_INTERVAL=10 → config["engine"]["poll_interval"] = 10
Strategy: try progressively joining remaining parts with '_' to match
actual config keys (handles multi-word keys like poll_interval).
"""
for key, value in os.environ.items():
if not key.startswith(prefix + "_"):
continue
parts = key[len(prefix) + 1:].lower().split("_")
_set_nested(config, parts, value)
return config
def _set_nested(config: dict, parts: list[str], value: str) -> bool:
"""Try to set a value in a nested dict using underscore-separated parts.
For parts ["engine", "poll", "interval"], tries:
config["engine_poll_interval"], config["engine"]["poll_interval"],
config["engine"]["poll"]["interval"]
"""
# Try joining all parts as a single key at this level
full_key = "_".join(parts)
if full_key in config and not isinstance(config[full_key], dict):
config[full_key] = _coerce(config[full_key], value)
return True
# Try splitting: first N parts as section key, rest as nested
for i in range(1, len(parts)):
section_key = "_".join(parts[:i])
if section_key in config and isinstance(config[section_key], dict):
if _set_nested(config[section_key], parts[i:], value):
return True
return False
def _coerce(old: Any, value: str) -> Any:
"""Coerce a string env var value to match the type of the existing value."""
if isinstance(old, bool):
return value.lower() in ("true", "1", "yes")
if isinstance(old, int):
return int(value)
if isinstance(old, float):
return float(value)
return value
def load_config(config_path: str | None = "auto") -> dict[str, Any]:
"""Load config with layered resolution: defaults → YAML → env vars."""
config = copy.deepcopy(DEFAULT_CONFIG)
# Load YAML if exists
if config_path == "auto":
config_path = str(Path.home() / ".catgo" / "config.yaml")
if config_path and Path(config_path).is_file():
try:
import yaml
with open(config_path) as f:
user_config = yaml.safe_load(f) or {}
config = _deep_merge(config, user_config)
except ImportError:
pass # yaml not installed, skip
except Exception:
pass # bad yaml, skip
# Apply environment variable overrides
config = _apply_env_vars(config)
return config
def get_default(config: dict, software: str, param: str) -> Any:
"""Get a default parameter value for a software type."""
defaults = config.get("defaults", {})
# Check software-specific defaults first
if software in defaults and param in defaults[software]:
return defaults[software][param]
# Check base software defaults (e.g., vasp_freq falls back to vasp)
base = software.split("_")[0]
if base in defaults and param in defaults[base]:
return defaults[base][param]
return None
def resolve_param(
param: str,
task_params: dict,
workflow_config: dict,
global_config: dict,
software: str,
) -> Any:
"""Resolve a parameter with 4-layer priority:
Task params > Workflow config > User config > System defaults.
"""
# 1. Task-level
if param in task_params:
return task_params[param]
# 2. Workflow-level
wf_defaults = workflow_config.get("defaults", {})
if software in wf_defaults and param in wf_defaults[software]:
return wf_defaults[software][param]
base = software.split("_")[0]
if base in wf_defaults and param in wf_defaults[base]:
return wf_defaults[base][param]
# 3+4. Global config (already merged with system defaults)
return get_default(global_config, software, param)
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| planning-and-task-breakdownaddyosmani/agent-skillsBreaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible. | Stars: 119k | Updated | Claude CodeCodex | Verified | Install |
| screenpipethedotmack/claude-memSearch screen and audio history recorded by Screenpipe and optionally save selected context into claude-mem. Use when the user asks to recall desktop activity, a meeting, or something seen in another app through Screenpipe. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| contact-widgetnexu-io/open-designSelf-contained floating chat widget with welcome screen, social links, meeting button, and message input. Single HTML file, zero dependencies. | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |
| meeting-notesnexu-io/open-designMeeting notes page — title bar with attendees, agenda checklist, decisions block, action items table with owners + dates, and a "next meeting" footer. Use when the brief mentions "meeting notes", "minutes", "1:1 notes", "all-hands recap", or "会议纪要". | Stars: 108k | Updated | Claude CodeCodex | Verified | Install |