# Cratis Chronicle CLI Operations > Inspect and operate a running Chronicle server with the cratis CLI - contexts and connection resolution, machine-readable output, failed partitions, observers, event sequences, read models, projections and jobs, and the authorization discipline for replay, retry, quarantine clearing and other mutations. Use when the question is about the state of a live store rather than source code. Do not use to diagnose a defect that reproduces locally in source. - Skill: `cratis/cratis-chronicle-cli-operations` (Agent Skill, multi-file: 2 files) - Install (CLI): `npx skillmds@latest add cratis/cratis-chronicle-cli-operations` - Raw SKILL.md: https://api.skillmd.com/api/skills/cratis/cratis-chronicle-cli-operations/raw - Safety review: pending - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: Coding & Dev Tools - License: MIT - Author: Cratis (https://skillmd.com/u/cratis) - Updated: 2026-09-17 - Page: https://skillmd.com/skills/cratis/cratis-chronicle-cli-operations --- # Operate a running Chronicle with the `cratis` CLI Source code says what *should* happen. When the question is what *is* happening in a live store — a projection that will not move, an observer that stopped, an event you are not sure was appended — read the server instead of the code. ## Verified product sources This skill is verified against these exact sources: | Package | Version | Purpose | | --- | --- | --- | | `Cratis.Cli` | `2.4.0` | the `cratis` command, its contexts, output formats, and command tree | | `Cratis.Chronicle` | `16.45.2` | the observer, partition, sequence, and job concepts the CLI reads | The CLI ships a **versioned machine-readable catalog of its own surface**, and that catalog is the authority for exact command names on the installed version: ```bash cratis llm-context # every command, option and argument as JSON cratis llm-context --schema # the JSON Schema for that document cratis --help # the same, one group at a time ``` Do not guess a command name from this file. This skill covers *when to reach for the CLI and how to read what comes back*; the catalog covers *what to type*. ## Install and set up `cratis` is a .NET tool published as `Cratis.Cli` with the command name `cratis`. Install it with `dotnet tool install -g Cratis.Cli`, or through Homebrew, or from a release binary. Follow with `cratis completions install`. Once per project, so every agent working it can reach the store: ```bash cratis init # detects the AI tools in use and writes project context cratis init --refresh # re-capture after upgrading the CLI ``` Two things worth knowing before running it: - The embedded catalog is a **snapshot**, not a live lookup. After a CLI upgrade it still describes the older surface; `cratis init` reports the mismatch and `--refresh` fixes it. - If the repository's instruction file is supplied by a shared package or a generated adapter, pass `--no-context`. Keep project-specific Chronicle connection guidance in a repository-owned project context file; never patch packaged or generated bytes. ## Reach the right server The connection string resolves in this exact order: 1. `--server ` 2. the `CHRONICLE_CONNECTION_STRING` environment variable 3. the active context's server in `~/.cratis/config.json` 4. `chronicle://localhost:35000` Credentials are composed separately, and only when the resolved string carries no embedded `user:pass@` and no `apiKey=`: a cached access token first, then the context's client id and secret. Prefer a named context over repeating a connection string: ```bash cratis context create --server chronicle://:35000 cratis context set cratis context show ``` A context also carries `--event-store` and `--namespace` defaults; both can be overridden per command with `-e` and `-n`. The event store defaults to `default` and the namespace to `Default`. **Be deliberate about which store you are pointed at.** The same commands read production and a local container, and several of them mutate. `cratis context show` before anything that writes; `--debug` prints the resolved config path, context, redacted connection string, and RPC timing to stderr. ## Choose the output format There is one output flag, `-o` / `--output`, with values `table`, `plain`, `json`, `json-compact`, and `auto`. **There is no `--json` flag.** - `auto` (the default) resolves to `json-compact` when an AI-agent environment is detected, `plain` when `NO_COLOR` is set, `json` when stdout is redirected, and `table` otherwise. - `-q` / `--quiet` prints identifiers only, one per line — what you want when piping one command into another. Combined with a JSON format it yields a JSON array of identifiers. - Prefer `plain` for large listings; JSON repeats every field name on every row. Prefer `json`/`json-compact` for `show`/`get` commands where you need the nested structure. Exit codes are stable and worth branching on: `0` success, `1` not found, `2` connection error, `3` server error, `4` authentication error, `5` validation error. ## Where to look | Question | Where to look | | --- | --- | | Is the server healthy at all? | `cratis chronicle diagnose` (exit code 3 when unhealthy) | | Why has this read model stopped updating? | `chronicle failed-partitions list --observer ` — the error is on the partition, not in the log | | An observer is quarantined — why? | `chronicle failed-partitions show --detailed`, with attempt history | | Did this event actually get appended? | `chronicle events get --event-type --event-source-id `, or `chronicle events tail` | | What is this event's shape in the store? | `chronicle event-types list`, then `chronicle event-types show ` | | Is this projection registered? | `chronicle projections list`, then `chronicle projections show ` | | What does the read model actually hold? | `chronicle read-models get `, or `read-models instances` | | Is a replay or migration still running? | `chronicle jobs list`, then `chronicle jobs get ` | | What does the server itself recommend? | `chronicle recommendations list` | Inspection commands are read-only: `diagnose`, `version`, `event-stores list`, `namespaces list`, `event-types list|show`, `events get|tail`, `observers list|show`, `failed-partitions list|show`, `projections list|show`, `read-models list|instances|get|snapshots|occurrences`, `jobs list|get`, `recommendations list`, `subscriptions list`, `identities list`, `users list`, `applications list`, and `auth status`. ## Read what comes back - **A failed partition does not retry itself.** It stays failed until something clears it, so a stale value is permanent rather than slow. That distinction is the whole diagnosis: "not arrived yet" and "will never arrive" look identical from the outside. - **Fix the cause before replaying.** Replaying into an unfixed handler fails the same way and buries the original error under a newer one. - **A quarantined observer does not resume by itself.** Clearing the quarantine is an explicit operation, and it is the last step, not the first. ## Before you change anything Default to read-only inspection. A request to diagnose a live store does not authorize replay, retry, removal, quarantine clearing, or another mutation. These commands mutate the running store: - `chronicle observers replay`, `replay-partition`, `retry-partition`, `clear-quarantine` - `chronicle jobs stop`, `chronicle jobs resume` - `chronicle recommendations perform`, `chronicle recommendations ignore` - `chronicle subscriptions add`, `chronicle subscriptions remove` - `chronicle users add`, `chronicle users remove` - `chronicle applications add`, `chronicle applications remove` Before any of them: 1. Name the exact server context, event store, namespace, observer, partition, event sequence, or job that will change. 2. Capture its exact pre-state and the failure evidence that justifies the operation. 3. Prepare the repository-approved inverse or safe compensation when one exists; stop when the effect cannot be safely reversed unless the owner separately authorizes that irreversible operation. 4. Show the dry run and obtain explicit authorization for that exact target and action. 5. Re-read the target immediately before mutation and stop on drift. ### Two confirmation traps **Confirmation prompts do not fire when stdout is not a terminal.** The CLI treats a non-interactive session as an implicit yes, so a destructive command run from a script, a pipeline, or an agent harness proceeds without `-y` and without asking. Never rely on the prompt as the safety mechanism — the authorization above is the safety mechanism. **Not every mutating command prompts, even interactively.** `jobs stop` and `jobs resume` describe themselves as prompting for confirmation, but no prompt exists in the implementation. `users add`, `applications add`, and `subscriptions add` do not prompt either. Reaching for `--yes` to silence a prompt you have not read is how the wrong store gets replayed. A failed partition you have not yet explained is not a thing to clear: read it, fix the handler, then replay. Clearing it first destroys the evidence and the same failure returns on the next event. ## What the CLI cannot do There is no `delete`, `purge`, `redact`, or `revise` command. Event redaction and revision are browser Workbench operations. Treat a request to "remove an event with the CLI" as a request that needs a different tool and a separate authorization. ## Related - The terminal Workbench (`cratis chronicle workbench`) presents the same read-only data as a navigable full-screen view; reach for it when you are exploring rather than answering one question. - The browser Workbench owns redaction, revision, and event-type authoring. - Source-level diagnosis belongs in the application's own diagnostic workflow; start there when the symptom reproduces locally, and start here when it is only visible on a running server.