Skills · Coding

Python Observability

Unverified30/40

Python observability patterns including structured logging, metrics, and distributed tracing. Use when adding logging, implementing metrics collection, setting up tracing, or debugging production systems.

Originally by wshobson · MIT

Claude CodePartialHas SKILL.md but declares no allowed-tools — Claude Code will ask for permission each time
CursorPartialPlain prose you can paste in — but no Cursor rules file
CodexPartialPlain prose you can paste in — but no AGENTS.md
Gemini CLIPartialPlain prose you can paste in
CopilotPartialPlain prose you can paste in — but no Copilot instructions file
npx agentalley add python-observability

This command does not work yet — the CLI is still being built. Until then, use Raw in the reader below to take the file.

Who is stuck, and on what

Python observability patterns including structured logging, metrics, and distributed tracing. Use when adding logging, implementing metrics collection, setting up tracing, or debugging production systems.

The whole source

No sign-in, no blur, nothing truncated
python-observability/SKILL.md230 lines6.9 KBRawView on GitHub
Frontmatter — 2 properties
namepython-observability
descriptionPython observability patterns including structured logging, metrics, and distributed tracing. Use when adding logging, implementing metrics collection, setting up tracing, or debugging production systems.
1---
2name: python-observability
3description: Python observability patterns including structured logging, metrics, and distributed tracing. Use when adding logging, implementing metrics collection, setting up tracing, or debugging production systems.
4---A5No allowed-tools declared — no way to tell what this skill may touch
5 
6# Python Observability
7 
8Instrument Python applications with structured logs, metrics, and traces. When something breaks in production, you need to answer "what, where, and why" without deploying new code.
9 
10## When to Use This Skill
11 
12- Adding structured logging to applications
13- Implementing metrics collection with Prometheus
14- Setting up distributed tracing across services
15- Propagating correlation IDs through request chains
16- Debugging production issues
17- Building observability dashboards
18 
19## Core Concepts
20 
21### 1. Structured Logging
22 
23Emit logs as JSON with consistent fields for production environments. Machine-readable logs enable powerful queries and alerts. For local development, consider human-readable formats.
24 
25### 2. The Four Golden Signals
26 
27Track latency, traffic, errors, and saturation for every service boundary.
28 
29### 3. Correlation IDs
30 
31Thread a unique ID through all logs and spans for a single request, enabling end-to-end tracing.
32 
33### 4. Bounded Cardinality
34 
35Keep metric label values bounded. Unbounded labels (like user IDs) explode storage costs.
36 
37## Quick Start
38 
39```python
40import structlog
41 
42structlog.configure(
43 processors=[
44 structlog.processors.TimeStamper(fmt="iso"),
45 structlog.processors.JSONRenderer(),
46 ],
47)
48 
49logger = structlog.get_logger()
50logger.info("Request processed", user_id="123", duration_ms=45)
51```
52 
53## Fundamental Patterns
54 
55### Pattern 1: Structured Logging with Structlog
56 
57Configure structlog for JSON output with consistent fields.
58 
59```python
60import logging
61import structlog
62 
63def configure_logging(log_level: str = "INFO") -> None:
64 """Configure structured logging for the application."""
65 structlog.configure(
66 processors=[
67 structlog.contextvars.merge_contextvars,
68 structlog.processors.add_log_level,
69 structlog.processors.TimeStamper(fmt="iso"),
70 structlog.processors.StackInfoRenderer(),
71 structlog.processors.format_exc_info,
72 structlog.processors.JSONRenderer(),
73 ],
74 wrapper_class=structlog.make_filtering_bound_logger(
75 getattr(logging, log_level.upper())
76 ),
77 context_class=dict,
78 logger_factory=structlog.PrintLoggerFactory(),
79 cache_logger_on_first_use=True,
80 )
81 
82# Initialize at application startup
83configure_logging("INFO")
84logger = structlog.get_logger()
85```
86 
87### Pattern 2: Consistent Log Fields
88 
89Every log entry should include standard fields for filtering and correlation.
90 
91```python
92import structlog
93from contextvars import ContextVar
94 
95# Store correlation ID in context
96correlation_id: ContextVar[str] = ContextVar("correlation_id", default="")
97 
98logger = structlog.get_logger()
99 
100def process_request(request: Request) -> Response:
101 """Process request with structured logging."""
102 logger.info(
103 "Request received",
104 correlation_id=correlation_id.get(),
105 method=request.method,
106 path=request.path,
107 user_id=request.user_id,
108 )
109 
110 try:
111 result = handle_request(request)
112 logger.info(
113 "Request completed",
114 correlation_id=correlation_id.get(),
115 status_code=200,
116 duration_ms=elapsed,
117 )
118 return result
119 except Exception as e:
120 logger.error(
121 "Request failed",
122 correlation_id=correlation_id.get(),
123 error_type=type(e).__name__,
124 error_message=str(e),
125 )
126 raise
127```
128 
129### Pattern 3: Semantic Log Levels
130 
131Use log levels consistently across the application.
132 
133| Level | Purpose | Examples |
134|-------|---------|----------|
135| `DEBUG` | Development diagnostics | Variable values, internal state |
136| `INFO` | Request lifecycle, operations | Request start/end, job completion |
137| `WARNING` | Recoverable anomalies | Retry attempts, fallback used |
138| `ERROR` | Failures needing attention | Exceptions, service unavailable |
139 
140```python
141# DEBUG: Detailed internal information
142logger.debug("Cache lookup", key=cache_key, hit=cache_hit)
143 
144# INFO: Normal operational events
145logger.info("Order created", order_id=order.id, total=order.total)
146 
147# WARNING: Abnormal but handled situations
148logger.warning(
149 "Rate limit approaching",
150 current_rate=950,
151 limit=1000,
152 reset_seconds=30,
153)
154 
155# ERROR: Failures requiring investigation
156logger.error(
157 "Payment processing failed",
158 order_id=order.id,
159 error=str(e),
160 payment_provider="stripe",
161)
162```
163 
164Never log expected behavior at `ERROR`. A user entering a wrong password is `INFO`, not `ERROR`.
165 
166### Pattern 4: Correlation ID Propagation
167 
168Generate a unique ID at ingress and thread it through all operations.
169 
170```python
171from contextvars import ContextVar
172import uuid
173import structlog
174 
175correlation_id: ContextVar[str] = ContextVar("correlation_id", default="")
176 
177def set_correlation_id(cid: str | None = None) -> str:
178 """Set correlation ID for current context."""
179 cid = cid or str(uuid.uuid4())
180 correlation_id.set(cid)
181 structlog.contextvars.bind_contextvars(correlation_id=cid)
182 return cid
183 
184# FastAPI middleware example
185from fastapi import Request
186 
187async def correlation_middleware(request: Request, call_next):
188 """Middleware to set and propagate correlation ID."""
189 # Use incoming header or generate new
190 cid = request.headers.get("X-Correlation-ID") or str(uuid.uuid4())
191 set_correlation_id(cid)
192 
193 response = await call_next(request)
194 response.headers["X-Correlation-ID"] = cid
195 return response
196```
197 
198Propagate to outbound requests:
199 
200```python
201import httpx
202 
203async def call_downstream_service(endpoint: str, data: dict) -> dict:
204 """Call downstream service with correlation ID."""
205 async with httpx.AsyncClient() as client:
206 response = await client.post(
207 endpoint,
208 json=data,
209 headers={"X-Correlation-ID": correlation_id.get()},
210 )
211 return response.json()
212```
213 
214## Detailed worked examples and patterns
215 
216Detailed sections (starting with `## Advanced Patterns`) live in `references/details.md`. Read that file when the navigation summary above is insufficient.
217 
218## Best Practices Summary
219 
2201. **Use structured logging** - JSON logs with consistent fields
2212. **Propagate correlation IDs** - Thread through all requests and logs
2223. **Track the four golden signals** - Latency, traffic, errors, saturation
2234. **Bound label cardinality** - Never use unbounded values as metric labels
2245. **Log at appropriate levels** - Don't cry wolf with ERROR
2256. **Include context** - User ID, request ID, operation name in logs
2267. **Use context managers** - Consistent timing and error handling
2278. **Separate concerns** - Observability code shouldn't pollute business logic
2289. **Test your observability** - Verify logs and metrics in integration tests
22910. **Set up alerts** - Metrics are useless without alerting
230 

Reviews

Installed this one?Write the first review and take the Trailblazer badge.

Reviews only open after a real install, so this is empty — and we leave it empty rather than invent one.

Alternatives

Also in Coding