Shell: Eżekuzzjoni tal-Kmandi
F'din il-paġna
Shell: Eżekuzzjoni tal-Kmandi
Din hija l-aktar għodda qawwija f’dan il-kors, u l-unika waħda fejn żball jista’ jdawwar il-qawwa tagħha kontra tiegħek. Għodod oħra jaqraw, jikkalkulaw, jew ifittxu; din taġixxi fuq magna. Dijanjosi tas-sistema mfassla ħażin tibqa’ sempliċiment dijanjosi; kmand ta’ tindifa mfassal ħażin iħassar fajls.
Din il-lezzjoni tkopri l-użu u l-miżuri ta’ sigurtà flimkien għax ma jistgħux jiġu separati. Ir-regola tad-deheb: il-mudell qatt m’għandu jeżegwixxi kmand li ma awtorizzajtx espliċitament minn qabel.
Kunċett u Arkitettura
Id-differenza mill-Code Interpreter hija kritika, u tinżel f’kelma waħda: iżolament. Il-Code Interpreter jaħdem f’sandbox li tista’ tintrema u iżolata min-netwerk fejn l-agħar xenarju possibbli huwa kalkolu żbaljat. Is-Shell taħdem f’ambjent li inti tikkonfigura—potenzjalment il-magna lokali tiegħek, potenzjalment server—fejn l-agħar xenarju possibbli huwa operazzjoni irreversibbli.
Din il-libertà hija l-għan: mingħajrha, id-distribuzzjoni u l-manutenzjoni awtomatizzata huma impossibbli. Madankollu, dan jeħtieġ li int tagħżel b’mod konxju l-ambitu tal-eżekuzzjoni. Kontenitur iddedikat b’privileġġi limitati huwa kważi dejjem l-għażla t-tajba meta mqabbel ma’ aċċess dirett għall-host.
from openai import OpenAI
client = OpenAI()
# The Shell tool is used via the Agents SDK
# because it requires a controlled execution environment
from agents import Agent, Runner
agent = Agent(
name="devops-assistant",
model="gpt-5.6-terra",
instructions=(
"You are a DevOps assistant. You execute system commands "
"to diagnose and resolve infrastructure issues. "
"Explain each command before executing it."
),
tools=[{"type": "shell"}]
)
Każ ta’ Użu: Dijanjosi tas-Sistema
# The agent can execute diagnostic commands
agent = Agent(
name="diagnostic",
model="gpt-5.6-terra",
instructions=(
"Diagnose the system state by executing appropriate commands. "
"Start with basic checks (disk, memory, CPU) "
"then dig deeper based on symptoms."
),
tools=[{"type": "shell"}]
)
# The agent will execute commands like:
# df -h -> disk space
# free -m -> memory
# top -bn1 -> CPU load
# systemctl status nginx -> service status
Sigurtà tal-Eżekuzzjoni
Tliet miżuri ta’ sigurtà jaħdmu flimkien, u l-ebda waħda mhix biżżejjed waħedha. L-allowlist hija l-fondazzjoni: intielenka dak li huwa permess minflok dak li huwa pprojbit, għax lista ta’ blokok (blocklist) dejjem titlef varjant—u biżżejjed wieħed biss. Timeout jipprevjeni kmand imblukkat milli jwaqqaf is-sistema. Direttorju tax-xogħol ristrett jillimita l-ħsara jekk kmand leġittimu jipponta lejn il-post żbaljat.
Ir-raba’ miżura, mhux teknika iżda tiswi daqs it-tlieta l-oħra flimkien: itlob konferma umana għal kwalunkwe operazzjoni distruttiva jew irreversibbli. Mudell jagħmel żbalji bl-istess kunfidenza meta jkollu raġun.
# Pattern: wrapper function with validation
import subprocess
import shlex
ALLOWED_COMMANDS = {
"diagnostic": ["df", "free", "top", "uptime", "ps", "netstat", "curl", "ping"],
"deployment": ["git", "docker", "npm", "pip"],
"files": ["ls", "cat", "head", "tail", "wc", "grep", "find"],
}
FORBIDDEN_COMMANDS = ["rm", "dd", "mkfs", "shutdown", "reboot", "kill", "passwd"]
def execute_secure_command(command: str, category: str = "diagnostic") -> dict:
"""Executes a command with security validation."""
# Parse the command
parts = shlex.split(command)
executable = parts[0]
# Check prohibitions
if executable in FORBIDDEN_COMMANDS:
return {
"error": f"Forbidden command: {executable}",
"reason": "This command is on the blocklist"
}
# Check permissions
allowed = ALLOWED_COMMANDS.get(category, [])
if executable not in allowed:
return {
"error": f"Command not allowed in category '{category}'",
"allowed": allowed
}
# Execute with timeout
try:
result = subprocess.run(
parts,
capture_output=True,
text=True,
timeout=30,
cwd="/tmp" # Restricted working directory
)
return {
"stdout": result.stdout,
"stderr": result.stderr,
"return_code": result.returncode
}
except subprocess.TimeoutExpired:
return {"error": "Timeout: command exceeded 30 seconds"}
Implimentazzjoni bħala Funzjoni Customizzata
L-implimentazzjoni ta’ dan int stess permezz tas-sejħa tal-funzjonijiet (function calling) toffri vantaġġ li m’għandekx tqisu bħala pjan ta’ riżerva: il-validazzjoni tgħix fil-kodiċi tiegħek. Inti tiddeċiedi l-allowlist, il-logging, u l-passi ta’ konferma, u tista’ tawditja kollox. Għal għodda b’dan il-livell ta’ riskju, li żżomm il-kontroll tal-punt ta’ kontroll huwa preferibbli, anke meta ssir disponibbli verżjoni nattiva.
tools = [
{
"type": "function",
"name": "execute_command",
"description": (
"Execute a shell command on the server. "
"Allowed commands: system diagnostics (df, free, top, uptime, ps), "
"file reading (ls, cat, head, tail, grep), "
"network (curl, ping, netstat). "
"Destructive commands are forbidden."
),
"parameters": {
"type": "object",
"properties": {
"command": {
"type": "string",
"description": "The shell command to execute"
},
"directory": {
"type": "string",
"description": "Working directory (default: /tmp)"
}
},
"required": ["command"],
"additionalProperties": False
},
"strict": True
}
]
import json
def handle_calls(response):
results = []
for item in response.output:
if item.type == "function_call" and item.name == "execute_command":
args = json.loads(item.arguments)
result = execute_secure_command(
args["command"],
category="diagnostic"
)
results.append({
"type": "function_call_output",
"call_id": item.call_id,
"output": json.dumps(result)
})
return results
# Full loop
response = client.responses.create(
model="gpt-5.6-terra",
input="Check disk space and available memory",
tools=tools
)
results = handle_calls(response)
final_response = client.responses.create(
model="gpt-5.6-terra",
input=response.output + results,
tools=tools
)
print(final_response.output_text)
Flussi ta’ Ħidma ta’ Awtomazzjoni
Distribuzzjoni Awtomatizzata
Id-distribuzzjoni (deployment) hija l-aktar każ ta’ użu prezzjuż, sakemm tifhem x’inhu qed tiddelega. Il-mudell ma jinventax il-proċedura: iħabbat flimkien passi li ddefinixxejt int, jivverifika outputs intermedjarji, u jieqaf fuq kwalunkwe anomaliija. Huwa operatur attenti, mhux dawk li jieħdu d-deċiżjonijiet.
tools_deploy = [
{
"type": "function",
"name": "execute_deploy_step",
"description": (
"Execute a deployment step. "
"Possible steps: git_pull, tests, build, deploy, rollback. "
"Each step is validated before proceeding to the next."
),
"parameters": {
"type": "object",
"properties": {
"step": {
"type": "string",
"enum": ["git_pull", "tests", "build", "deploy", "rollback"]
},
"parameters": {
"type": "object",
"properties": {
"branch": {"type": "string"},
"environment": {"type": "string", "enum": ["staging", "production"]}
}
}
},
"required": ["step"],
"additionalProperties": False
},
"strict": True
}
]
def execute_step(step: str, params: dict) -> dict:
scripts = {
"git_pull": "cd /app && git pull origin {branch}",
"tests": "cd /app && npm test",
"build": "cd /app && npm run build",
"deploy": "cd /app && ./deploy.sh {environment}",
"rollback": "cd /app && ./rollback.sh {environment}"
}
command = scripts[step].format(**params)
return execute_secure_command(command, "deployment")
Monitoraġġ u Allerti
Il-monitoraġġ jibdel il-loġika tal-bqija ta’ din il-lezzjoni: il-kmandi hawnhekk huma biss għall-qari, allura huma innocwi, u l-valur jinsab fl-interpretazzjoni. Mudell li jaqra l-ispazju tad-diska, it-tagħbija, u l-logs tal-iżbalji jinduna b’korrelazzjonijiet li limitu iżolat jitlifhom—u dan jagħmilha l-aktar punt ta’ dħul sikur biex tibda tuża din l-għodda.
def check_system_health() -> dict:
"""Checks system health and returns a report."""
checks = {}
# Disk space
r = execute_secure_command("df -h /", "diagnostic")
checks["disk"] = r
# Memory
r = execute_secure_command("free -m", "diagnostic")
checks["memory"] = r
# Services
for service in ["nginx", "postgresql", "redis"]:
r = execute_secure_command(
f"systemctl is-active {service}",
"diagnostic"
)
checks[f"service_{service}"] = r
return checks
# The AI analyzes the results and suggests actions
response = client.responses.create(
model="gpt-5.6-terra",
input=f"Analyze this system health report and identify issues:\n"
f"{json.dumps(check_system_health(), indent=2)}",
tools=tools
)
Punti Ewlenin
- L-għodda Shell tgħaqqad il-mudell mal-infrastruttura tiegħek
- Implimenta allowlists u blocklists għall-kmandi
- Uża timeouts u direttorji tax-xogħol ristretti
- Is-sejħa tal-funzjonijiet (function calling) tippermettilek timplimenta s-Shell bħala għodda customizzata
- Għaqqad is-Shell ma’ għodod oħra għal flussi ta’ ħidma DevOps sħaħ