* Codex auto mode: grant the orx data dir as a sandbox writable root workspace-write denies writes outside the workspace, and the orx SQLite store lives under ~/.local/share (WAL + read-write open), so every store-touching orx command died with 'unable to open database file' under Auto mode. Add store::data_dir() to the turn's writable_roots alongside the shared git dir. * Review fixes: pin ORX_DATA_DIR to the granted root, ensure_ naming - Pin ORX_DATA_DIR on the codex child (after prepare_env) so the store the sandboxed orx resolves is the exact path granted as a writable root — synced env or a relative override could otherwise diverge. - Rename orx_data_dir -> ensure_orx_data_dir (codebase ensure_ prefix for create-if-missing helpers); note the grant is dir-wide. - Dedent the '-c replaces user roots' note: it covers both roots. * Review round 2: match-expression pin, stale options() doc, asymmetry note
OpenResearch CLI (orx)
A small command-line interface for OpenResearch. Log in via the browser, then query your projects from the terminal.
Requirements
-
Rust (stable) with Cargo. Install via rustup:
curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
Build & run
# Build (debug):
cargo build
# Run during development:
cargo run -- login
cargo run -- projects
# Build an optimized release binary at target/release/orx:
cargo build --release
./target/release/orx login
# Or install `orx` onto your PATH (~/.cargo/bin):
cargo install --path .
orx login
orx projects
Run the tests with cargo test.
Commands
| Command | Description |
|---|---|
orx login [--api-url <url>] |
Opens your browser, authenticates, and stores a personal access token. |
orx logout |
Removes the stored token. |
orx install-skills [--agent claude|codex|opencode|cursor|all] |
Installs the orx skill shim into local coding agents (Claude Code, Codex, OpenCode, Cursor) so they auto-discover the CLI (orx login also offers this). |
orx projects [--all] |
Lists your projects, grouped by organization. --all includes archived. |
orx compute [--gpu <id>] [--count <n>] [--provider <name>] | --cpu] |
Lists the GPU compute catalog across all providers and regions, or CPU-only offers with --cpu, sorted by price. |
orx instance create <orgId> (--gpu <id> [--count <n>] [--disk <gb>] [--provider <name>] | --cpu <flavor> [--vcpus <n>]) |
Spins up a standalone instance in an organization (not tied to an experiment), like the dashboard's "Spin up". |
orx exp status <expId> |
Shows an experiment's status, branch, parent, run command, and latest run, plus a local git diff recipe for what the run changed. |
orx exp cmd <expId> [--set <command>] |
Views or sets the experiment's run command. |
orx exp run <expId> (--gpu <id> [--count <n>] [--disk <gb>] [--provider <name>] | --cpu <flavor> [--vcpus <n>] | --sandbox <id>) |
Launches a run on new GPU, new CPU-only, or existing compute. |
orx exp cancel <expId> |
Cancels the in-flight run. |
orx lit "<query>" [--limit <n>] [--json] |
Full-text search alphaXiv's paper corpus (no login required). |
orx paper <id|url> [--full] |
Fetch a paper's machine-readable report, or its full text with --full. |
orx version [--check] [--json] |
Prints the CLI version; --check compares it to the latest release. |
orx update [--dry-run] [--force] |
Updates orx by re-running the release installer (installer installs only). |
Updating
orx update only manages binaries installed by the release installer script
(it checks the cargo-dist install receipt at
${XDG_CONFIG_HOME:-~/.config}/openresearch-cli/). If you installed with
cargo install, update the same way instead. In interactive terminals orx
also checks for new releases in the background at most once a day and prints
a one-line notice to stderr; set ORX_NO_UPDATE_CHECK=1 (or
OPENRESEARCH_CLI_DISABLE_UPDATE=1) to turn that off. Agents, scripts, pipes,
and CI never see the notice.
Configuration
-
API URL resolves from
--api-url→OPENRESEARCH_API_URL→ the built-in default (http://localhost:4000). Point it at the production API host for real use:export OPENRESEARCH_API_URL=https://api.openresearch.sh -
Credentials are stored at
${XDG_CONFIG_HOME:-~/.config}/openresearch/credentials.json(mode0600).
How login works
orx login starts a temporary HTTP listener on a random 127.0.0.1 port, opens
{api}/auth/cli/login in your browser, and waits for the API to redirect back
with a freshly minted personal access token after you authenticate. The token is
sent as Authorization: Bearer … on every subsequent request.