FastAPI AnyIO TaskGroup Cancellation Shielding for Critical Cleanup and Audit Pipelines

--- description: FastAPI AnyIO TaskGroup Cancellation Shielding for Critical Cleanup and Audit Pipelines - Enforces deterministic cancellation shielding and detached lifecycle execution for critical post-transaction steps, telemetry, and audit logging in FastAPI applications powered by AnyIO. Prevents silent data corruption, orphaned lock states, and unrecorded compliance trails when HTTP clients disconnect or proxy timeouts propagate cancellation cascades into coroutine trees. globs: app/**/*.py, src/**/*.py, routers/**/*.py, services/**/*.py alwaysApply: true verified_by: "agentrules.org" license: "CC-BY-4.0" source: "https://agentrules.org/rules/fastapi-anyio-cancellation-shielding-standard-fastapi-anyio-ta-964" --- # FastAPI AnyIO TaskGroup Cancellation Shielding for Critical Cleanup and Audit Pipelines > **Severity**: CRITICAL | **Category**: backend > **Anti-Pattern Warning**: Unshielded post-transaction cleanup and audit coroutines aborted mid-execution during client disconnects ## System Constraints You are a Principal Backend Architect enforcing strict concurrency and cancellation semantics in FastAPI services powered by AnyIO and Starlette. Mandate explicit cancellation shielding on all critical side-effects, audit trails, and post-transaction cleanups that must outlive client-aborted HTTP request scopes. Prohibit raw unshielded await calls inside request-bound finally blocks and ban naked asyncio.shield usage in favor of structured AnyIO cancellation scopes. ## Strict Guidelines - Engineers MUST ALWAYS encapsulate mission-critical audit logging, billing reconciliation, and external telemetry inside `with anyio.CancelScope(shield=True):` when executed in request contexts. - Engineers MUST NOT rely on unshielded `finally` blocks for transactional state cleanups (e.g., releasing distributed Redis locks, invalidating active sessions) because `CancelledError` will terminate subsequent `await` calls immediately. - Engineers MUST NOT use naked `asyncio.shield()` directly in FastAPI routes, as it fails to properly tie task lifecycles to structured AnyIO task groups and leaks detached tasks on process shutdown. - Code SHOULD offload non-blocking, non-critical background jobs to an application-lifespan scoped AnyIO `TaskGroup` rather than relying on endpoint-local `BackgroundTasks` that terminate unexecuted if cancellation strikes before response dispatch. - Services MUST ensure database rollbacks and compensating transactions run within an explicit AnyIO shielded scope if executed inside an async exception handler triggered by `anyio.get_cancelled_exc_class()`. ## Bad vs Good Implementations ### Unshielded Audit Logging & Cleanup vs Shielded AnyIO CancelScope #### Bad Antipattern (AI Common Hallucination) > **Why it fails**: When an HTTP client aborts the request or proxy times out, Starlette/Uvicorn injects a cancellation token into the endpoint coroutine. Any `await` inside the `except` or `finally` block raises `CancelledError` (or `anyio.get_cancelled_exc_class()`), immediately aborting the refund and dropping critical compliance audit logs silently. ```python import anyio from fastapi import APIRouter, Depends, HTTPException from app.services.payment import PaymentGateway from app.services.audit import AuditLogger router = APIRouter() @router.post("/checkout") async def checkout(order_id: str, payment: PaymentGateway = Depends(), audit: AuditLogger = Depends()): tx = await payment.charge(order_id) try: await payment.verify_receipt(tx.id) except Exception as exc: # FLAW: If client disconnects here, cancellation raises CancelledError # during the await, aborting refund and audit trails midway! await payment.reverse_charge(tx.id) await audit.record_failure(order_id, str(exc)) raise HTTPException(status_code=500, detail="Processing failed") finally: # FLAW: In an HTTP disconnect cascade, this await is cancelled immediately await audit.record_access(order_id) ``` #### Production Standard (Correct) > **Standard**: Wrapping compensating transactions and cleanup logic in `with anyio.CancelScope(shield=True):` prevents incoming cancellation signals from interrupting asynchronous I/O operations, ensuring audit logs and monetary reversals complete reliably. ```python import anyio from fastapi import APIRouter, Depends, HTTPException from app.services.payment import PaymentGateway from app.services.audit import AuditLogger router = APIRouter() @router.post("/checkout") async def checkout(order_id: str, payment: PaymentGateway = Depends(), audit: AuditLogger = Depends()): tx = await payment.charge(order_id) try: await payment.verify_receipt(tx.id) except Exception as exc: # Invariant: Compensating transaction and failure logging are shielded # from client disconnect cancellation cascades. with anyio.CancelScope(shield=True): await payment.reverse_charge(tx.id) await audit.record_failure(order_id, str(exc)) raise HTTPException(status_code=500, detail="Processing failed") finally: # Invariant: Access trail and resource release complete deterministically with anyio.CancelScope(shield=True): await audit.record_access(order_id) ```

Architectural Specification & Invariant Rationale

Architectural Invariant Design: FastAPI AnyIO TaskGroup Cancellation Shielding for Critical Cleanup and Audit Pipelines

FastAPI AnyIO TaskGroup Cancellation Shielding for Critical Cleanup and Audit Pipelines defines an industrial-grade behavioral standard for backend engineering. Enforces deterministic cancellation shielding and detached lifecycle execution for critical post-transaction steps, telemetry, and audit logging in FastAPI applications powered by AnyIO. Prevents silent data corruption, orphaned lock states, and unrecorded compliance trails when HTTP clients disconnect or proxy timeouts propagate cancellation cascades into coroutine trees. Modern software architectures require strict adherence to deterministic patterns to guarantee high availability, maintainability, and zero security vulnerabilities.

Framework Paradigm Shift & Drift MechanicsLLMs are trained on millions of open-source repositories containing outdated code, legacy idioms, and unpatched antipatterns. When coding assistants generate implementations for this stack, they default to common statistical tokens: Unshielded post-transaction cleanup and audit coroutines aborted mid-execution during client disconnects Without an explicit system instruction file, developers spend hours debugging silent failures.

Failure Modes & Production Blast Radius

Hazard Analysis: What Breaks Without This Rule

The following failure modes have been cataloged from hundreds of hours of production code reviews where AI assistants produced non-compliant implementations:

Failure Mode	AI Symptom (The Anti-Pattern)	Severity	Production Impact & Remedy
AI Hallucination & Deprecation Drift	Unshielded post-transaction cleanup and audit coroutines aborted mid-execution during client disconnects	Critical	Violating this industrial standard introduces regression bugs, security risks, and unexpected runtime failures in production. Enforce explicit behavioral system directives across all AI coding assistants in your repository.
Silent Architecture Inconsistency	AI assistants produce mixed paradigms across files, creating fragmented codebase conventions.	Medium	Increases cognitive load during code reviews and leads to maintenance debt across engineering teams. Pin project conventions in .cursor/rules, CLAUDE.md, and AGENTS.md so all team agents follow identical guardrails.

Architectural Directives in Detail

Step-by-Step Production Directives

Each rule constraint below is injected into the AI agent's system prompt to establish deterministic boundaries:

Engineers MUST ALWAYS encapsulate mission-critical audit logging, billing reconciliation, and external telemetry inside `with anyio.CancelScope(shield=True)

Engineers MUST ALWAYS encapsulate mission-critical audit logging, billing reconciliation, and external telemetry inside `with anyio.CancelScope(shield=True):` when executed in request contexts.

Rationale:AI models frequently generate unconstrained code here because statistical language distributions lack project-specific architectural context. Mandating this directive ensures that generated code adheres to verified industrial production requirements.