Operator-ready prompt for reuse, tuning, and workspace runs.
This item is set up for developers who want to inspect the original language, fork it into Workspace, and adapt the evidence model without losing the source prompt structure.
Implementation handoffs, eval setup, and prompt tuning where you need the original structure intact.
Inspect first, copy once, then fork into Workspace when you want variants, notes, and model settings attached to the same run.
Swap domain facts, examples, and any hard-coded entities for your own context.
Tighten the evidence or verification requirement if this is headed toward production.
Decide which failure mode you want to evaluate first before you branch the prompt.
This prompt already carries implementation detail, tool context, and a final-output instruction. Keep that structure intact when you tune it, or your comparison runs get noisy fast.
Open this prompt inside Workspace when you want a live iteration loop.
Copy for quick reuse, or run it in Workspace to keep prompt variants, model settings, and prompt-history changes in one place.
Structured source with 32 active lines to adapt.
Already linked to a challenge workflow.
Sign in to keep private prompt variations.
Prompt content
Original prompt text with formatting preserved for inspection and clean copy.
Set up a basic Claude agent using the Claude Agents SDK. Define two simple tools: `get_instance_metadata(instance_id: str)` returning instance tags and `threat_intel_lookup(hash: str)` returning known threats. Configure the agent to use Claude Opus 4.1. Write a test case where the agent receives a high CPU alert and needs to use these tools to gather context.
```python
# agent_main.py
from anthropic import Anthropic
from anthropic_agents import Agent, tool
anthropic_client = Anthropic()
@tool
def get_instance_metadata(instance_id: str) -> str:
"""Gets metadata like tags for a given cloud instance ID."""
# Simulate API call
if instance_id == "i-abcdefg123":
return "tags: prod, web-server"
return "No metadata found."
@tool
def threat_intel_lookup(hash: str) -> str:
"""Looks up a file hash in threat intelligence databases."""
# Simulate API call
if hash == "malicious_process.sh_hash":
return "Known cryptocurrency miner identified."
return "No threat found for this hash."
agent = Agent(
client=anthropic_client,
model="claude-3-opus-20240229", # Or newer model if available
tools=[get_instance_metadata, threat_intel_lookup],
system_prompt="You are a cloud security analyst. Analyze alerts, use tools, and provide classification and remediation."
)
async def run_triage(alert: str):
response = await agent.run(alert)
print(response.content)
# Example usage:
# asyncio.run(run_triage("High CPU alert on instance i-abcdefg123. Suspicious process 'malicious_process.sh' detected."))
```Adaptation plan
Keep the source stable, then branch your edits in a predictable order so the next prompt run is easier to evaluate.
Hold the task contract and output shape stable so generated implementations remain comparable.
Update libraries, interfaces, and environment assumptions to match the stack you actually run.
Test failure handling, edge cases, and any code paths that depend on hidden context or secrets.
Copy once for a pristine source snapshot, then move the prompt into Workspace when you want variants, run history, and side-by-side tuning without losing the original.
Prompt diagnostics
Quick signals for how structured this prompt already is and where adaptation work is likely to happen first.
This prompt already mixes executable detail with instructions, so the safest path is to tune examples and interfaces before you rewrite the overall scaffold.
Autonomous Cloud Security Triage Agent
This challenge tasks you with developing an autonomous cloud security triage agent. Utilizing the Claude Agents SDK, you will build an intelligent agent capable of analyzing incoming security alerts from various cloud environments, distinguishing between false positives and genuine threats, and providing detailed explanations and remediation recommendations. The agent will employ Claude Opus 4.1's advanced extended thinking capabilities to reason through complex alert data, correlate information across multiple sources, and leverage specialized tools served by TorchServe for deeper analysis (e.g., malware analysis, anomaly detection). The solution requires robust integration with monitoring systems to ingest alerts and generate actionable insights, significantly reducing the burden on human security teams by automating the initial, often time-consuming, triage process. The agent must be capable of explaining its reasoning process to human analysts, fostering trust and transparency.
Use the challenge page to recover the original task boundaries before you tune the prompt. That keeps your variants grounded in the same evaluation target instead of drifting into a different problem.