Operations

Open this section when you need to run index work repeatably: install it, exercise retrieval flows, diagnose replay drift, release it, or recover from failure without guessing what the package expected.

Operating Loop

flowchart LR
    setup["setup"]
    query["run retrieval flow"]
    inspect["inspect replay output"]
    recover["recover from drift"]
    release["release package"]
    proof["tests and artifacts"]

    setup --> query --> inspect --> recover --> release
    query --> proof
    inspect --> proof

Index operations need to make replayability practical, not rhetorical. A maintainer should be able to move from a clean setup to a reproducible query path, inspect the replay evidence, and recover from drift without treating the backend as a black box.

Read These First

• open Installation and Setup first when you need a clean package starting point
• open Observability and Diagnostics when retrieval or replay behavior no longer matches expectation
• open Failure Recovery when index output or search behavior has already gone wrong

Operational Risk

The main operational risk here is hiding retrieval and replay assumptions in environment state or one-off debugging habits.

First Proof Check

• pyproject.toml, README.md, and package-local entrypoints for checked-in operating truth
• tests and runnable workflows for evidence that the package can be operated repeatably
• release notes and version metadata when the work changes caller expectations

Pages In This Section

• Installation and Setup
• Local Development
• Common Workflows
• Observability and Diagnostics
• Performance and Scaling
• Failure Recovery
• Release and Versioning
• Security and Safety
• Deployment Boundaries

Leave This Section When

• leave for Interfaces when the live problem is contract shape rather than package operation
• leave for Architecture when a workflow problem exposes structural drift underneath it
• leave for Quality when the package runs but the real question is whether the evidence is strong enough

Design Pressure

If replay drift can only be diagnosed by backend intuition or private habits, the package is not operationally honest enough yet. The checked-in workflow needs to expose how search behavior is examined and recovered.