# Watcher > Documentation for Watcher, Apollo Research's coding-agent monitoring platform: > it records what agents do (sessions and trajectories), grades that activity > with LLM graders, can intervene in real time, and gives teams a dashboard to > review what happened. Every page is available as raw markdown at its URL plus `.md`. Found an issue in these docs? Report it: https://github.com/ApolloResearch/watcher-bin/issues/new?title=Docs%20feedback ## Before upgrading Watcher Read the [Releases index](https://watcher-docs.apolloresearch.ai/releases/index.md) before every upgrade. Determine the installed and target versions, then read every release page and linked full changelog with a version greater than the installed version and no greater than the target. Include versions you skip. Read releases even when they list no required actions, to understand behavior and compatibility changes. Apply required steps oldest first, using the self-hosted and Watcher Cloud audience labels. Use [Update Watcher](https://watcher-docs.apolloresearch.ai/self-hosting/updates.md) for the procedure and [Verify an update](https://watcher-docs.apolloresearch.ai/self-hosting/update-verification.md) for self-hosted verification. ## Docs - [Gateway hooks](https://watcher-docs.apolloresearch.ai/api/hooks.md): Connect gateways to Watcher for enforcement, observation, or transcript ingestion, including Aperture authentication and delivery limits. - [API](https://watcher-docs.apolloresearch.ai/api/index.md): The Watcher HTTP API for programmatic access: base URL, authentication with organization API keys, browsing the endpoints, and the Python SDK binding. - [Monitor customization](https://watcher-docs.apolloresearch.ai/api/monitor-customization.md): Read, replace, and clear organization instructions for managed live monitors through the Watcher API. - [Python SDK](https://watcher-docs.apolloresearch.ai/api/python-sdk.md): The typed Python client for the Watcher API: wheel and source installation, configuration, ingest, grading, sessions, trajectories, pagination, and error handling. - [Containers](https://watcher-docs.apolloresearch.ai/client/containers.md): Run the Watcher client inside containers with coding-agent hooks, authentication, and persistent settings. - [External credential helpers](https://watcher-docs.apolloresearch.ai/client/credential-helpers.md): Configure a trusted local CLI to supply renewable bearer tokens to a self-hosted Watcher authentication gateway. - [Hardening dev containers](https://watcher-docs.apolloresearch.ai/client/devcontainer-hardening.md): Security considerations and design options for isolating Watcher from an untrusted development container. - [Dev containers](https://watcher-docs.apolloresearch.ai/client/devcontainer.md): Running the Watcher client in a VS Code dev container, reusing the developer's existing sign-in when available. - [Signing in without a browser](https://watcher-docs.apolloresearch.ai/client/headless-sign-in.md): Authenticate Watcher on headless machines with device-code sign-in, organization API keys, external credential helpers, or shared container credentials. - [Client](https://watcher-docs.apolloresearch.ai/client/index.md): Install and update the Watcher client, configure coding-agent hooks and settings, and diagnose problems with watcher doctor and watcher logs. - [Excluding a directory with .watcher-ignore](https://watcher-docs.apolloresearch.ai/client/watcher-ignore.md): How a .watcher-ignore file keeps one directory's coding-agent sessions off your Watcher deployment, and what it leaves unchanged. - [Analyzer](https://watcher-docs.apolloresearch.ai/concepts/analyzer.md): Use Analyzer to review coding-agent sessions, inspect security decisions, and manage Watcher monitors and organization settings. - [Blocking and trailing review](https://watcher-docs.apolloresearch.ai/concepts/blocking-and-trailing-review.md): How Watcher reviews a live session: blocking review gates tool calls before they run, trailing review grades the trajectory as the session progresses. - [Graders](https://watcher-docs.apolloresearch.ai/concepts/graders.md): What a Watcher grader is: a versioned LLM grading rubric with score dimensions, the built-in graders, where they run, and how to create your own. - [Concepts](https://watcher-docs.apolloresearch.ai/concepts/index.md): The Watcher vocabulary: sessions, trajectories, graders, blocking vs. trailing review, and organization settings. - [Organization settings](https://watcher-docs.apolloresearch.ai/concepts/organization-settings.md): Configure Watcher organization settings, data retention, client policies, and Slack notifications in Analyzer. - [Sessions and trajectories](https://watcher-docs.apolloresearch.ai/concepts/sessions-and-trajectories.md): How Watcher organizes recorded coding-agent activity: sessions, the trajectories inside them, and which messages a grade applies to. - [Subagent coverage](https://watcher-docs.apolloresearch.ai/concepts/subagent-coverage.md): How Watcher reviews subagent tool calls through agent hooks, with a Claude Code example and limits on conversation visibility. - [Set up GitHub Copilot CLI](https://watcher-docs.apolloresearch.ai/integrations/copilot.md): Install Copilot recording and blocking review for every Copilot session, or for one project, with the Watcher SDK and without the local Watcher app. - [Integrations](https://watcher-docs.apolloresearch.ai/integrations/index.md): Choose a Watcher connection for coding agents, frameworks, gateways, compliance data and observability tools. - [Automate Analyzer updates](https://watcher-docs.apolloresearch.ai/self-hosting/automatic-updates.md): Set up approved automatic updates for a self-hosted Watcher deployment using a host scheduler, deployment pipeline, or internal mirror. - [AWS infrastructure](https://watcher-docs.apolloresearch.ai/self-hosting/aws-infrastructure.md): Provision self-hosted Watcher on AWS with the bundled Terraform modules. - [Client deployment](https://watcher-docs.apolloresearch.ai/self-hosting/client-deployment.md): Install the Watcher client on individual machines, shared VM images, or managed endpoints. - [Deployment example](https://watcher-docs.apolloresearch.ai/self-hosting/deployment-example.md): Compare the authentication paths into a self-hosted Watcher deployment. - [EKS reference](https://watcher-docs.apolloresearch.ai/self-hosting/eks-reference.md): Adapt the supplied EKS Terraform reference with private cluster access, RDS, an existing ALB, and staged Helm installation. - [Gateway deployment reference](https://watcher-docs.apolloresearch.ai/self-hosting/gateway-deployments.md): Reference diagrams showing how Tailscale Aperture and LiteLLM connect coding agents, model providers, and Watcher for review or tool blocking. - [Self-hosting](https://watcher-docs.apolloresearch.ai/self-hosting/index.md): Choose the self-hosted Watcher deployment, client rollout, security, and operations guidance for your environment. - [Kubernetes infrastructure](https://watcher-docs.apolloresearch.ai/self-hosting/kubernetes-infrastructure.md): Deploy self-hosted Watcher into an existing Kubernetes cluster with a Helm example, architecture diagram, and verification steps. - [MDM rollout](https://watcher-docs.apolloresearch.ai/self-hosting/mdm.md): Roll out the Watcher client with an MDM, protect agent hook configuration, and manage startup and settings. - [Operations](https://watcher-docs.apolloresearch.ai/self-hosting/operations.md): Back up, restore, audit, and collect diagnostics from a self-hosted Watcher deployment. - [Watcher Self-Host: Create or Customize a Grader](https://watcher-docs.apolloresearch.ai/self-hosting/prompts/custom-grader.md): The agent runbook for creating or customizing an LLM grader on a self-hosted Watcher deployment, and rolling it out for live blocking or on-demand grading. - [Watcher Self-Host: Infrastructure Prerequisites](https://watcher-docs.apolloresearch.ai/self-hosting/prompts/prerequisites.md): The agent runbook that provisions self-hosting infrastructure (Docker host, network, database, TLS) before guided setup; AWS is the reference implementation. - [Watcher Self-Host: Guided Setup](https://watcher-docs.apolloresearch.ai/self-hosting/prompts/setup.md): The agent runbook for installing self-hosted Watcher and arranging ongoing updates on the Docker host. - [Iru provider sample](https://watcher-docs.apolloresearch.ai/self-hosting/provider-samples/iru.md): Downloadable Iru scripts for installing, auditing, repairing, and removing the Watcher client on managed Macs. - [Security](https://watcher-docs.apolloresearch.ai/self-hosting/security.md): Review self-hosted Watcher data collection, storage, network connections, telemetry, auditing, and retention. - [Setup](https://watcher-docs.apolloresearch.ai/self-hosting/setup.md): Install self-hosted Watcher, then verify an existing update job or record an explicit automation deferral. - [Supply chain security](https://watcher-docs.apolloresearch.ai/self-hosting/supply-chain.md): Control the client and server artifacts used by a self-hosted Watcher deployment. - [Verify an update](https://watcher-docs.apolloresearch.ai/self-hosting/update-verification.md): Verify a self-hosted Watcher server update on one test machine before updating the client fleet. - [Update Watcher](https://watcher-docs.apolloresearch.ai/self-hosting/updates.md): Check release-specific upgrade steps, update a self-hosted Watcher server, and update clients on self-hosted or cloud deployments. - [Supported agents](https://watcher-docs.apolloresearch.ai/supported-agents.md): Supported local agents, Aperture gateway routes, and current Claude Tag limits. ## Releases - [Release page catalog](https://watcher-docs.apolloresearch.ai/releases/llms.txt): Every release version, raw Markdown page, affected upgrade audience, and public full changelog.