Command reference¶
All 124 commands. Every read command accepts --json; exit codes are 0 ok, 1 failures/findings, 2 usage/config, 3 guard refusal.
Core & flat commands¶
| Command | What it does |
|---|---|
qont ci |
Run the suite and emit a markdown report (regressions + hardware cost). |
qont completion |
Show how to enable shell completion. |
qont cost |
Estimate what running the discovered tests on real hardware would cost. |
qont dashboard |
Render run history as a self-contained HTML dashboard. |
qont diff |
Semantic diff of two circuits (register names and formatting ignored). |
qont doctor |
Diagnose the environment: imports, extras, config, catalog freshness. |
qont explain |
Failure forensics: which outcomes drove a failing check, and why. |
qont fuzz |
Flakiness hunter: rerun the suite under a seed sweep, estimate pass rates. |
qont hash |
Print a circuit's canonical content hash. |
qont health |
Provider health: per-device reliability from calibration and local history. |
qont init |
Scaffold a Qontinuum project: example test, config, gitignore entries. |
qont mock |
Generate synthetic measurement counts (test fixtures, demos, docs). |
qont plan |
Produce a concrete execution plan: device, estimates, risks, fallbacks. |
qont recommend |
Rank hardware for the discovered workload and explain the top choice. |
qont route |
Recommend which hardware should run the discovered tests. |
qont run |
Run the discovered tests on REAL hardware, guarded by a spend cap. |
qont test |
Discover and run quantum tests (files matching q_test_*.py). |
qont version |
Print the qontinuum version. |
See Execution intelligence for recommend, plan,
and health.
qont bench¶
Mirror, RB-lite, and QV-style benchmarks (estimates, labeled as such).
| Command | What it does |
|---|---|
qont bench compare |
Head-to-head: which backend treats your circuits better? |
qont bench mirror |
Mirror-circuit benchmark: run U·U†, measure survival of |
qont bench qv |
Quantum-volume-style estimate (heavy-output probability per width). |
qont bench rb |
RB-lite: exponential decay fit over mirrored random-layer circuits. |
qont bench suite |
The standard benchmark battery across several backends. |
qont budget¶
Hardware spend budgets and the run ledger.
| Command | What it does |
|---|---|
qont budget forecast |
Project this month's spend from the current run cadence. |
qont budget ledger |
Recorded hardware-run spend entries. |
qont budget reset |
Clear the spend ledger (caps stay configured). |
qont budget set |
Set a spend cap that qont run enforces before submitting anything. |
qont budget show |
Current caps. |
qont budget status |
Spend vs caps, from the run ledger. |
qont cache¶
Simulation result cache (seeded runs only).
| Command | What it does |
|---|---|
qont cache clear |
Delete every cached result. |
qont cache gc |
Garbage-collect old cache entries. |
qont cache ls |
List cached results. |
qont cache stats |
Cache size and entry count. |
qont circuit¶
Inspect, convert, and generate circuits.
| Command | What it does |
|---|---|
qont circuit canon |
Print the versioned canonical form (what the content hash covers). |
qont circuit convert |
Convert between OpenQASM dialects (also accepts Cirq/PennyLane via API). |
qont circuit diff |
Semantic diff (same engine as top-level qont diff). |
qont circuit equiv |
True unitary equivalence via statevector (up to global phase, <=12 qubits). |
qont circuit hash |
Print the circuit's canonical content hash. |
qont circuit mirror |
Emit the mirror-benchmark circuit U·U†: ideal outcome is all zeros. |
qont circuit random |
Generate a reproducible random circuit (fuzzing / benchmarking input). |
qont circuit show |
Draw the circuit as unicode art. |
qont circuit stats |
Gate counts, depth, SPAM profile, and content hash. |
qont circuit transpile |
Show what the device will actually run: depth/gate deltas after targeting. |
qont config¶
Project configuration (.qontinuum/config.toml).
| Command | What it does |
|---|---|
qont config edit |
Open the config file in $EDITOR. |
qont config get |
Print one config value (built-in default when unset). |
qont config list |
Show every known key: effective value, source, and description. |
qont config path |
Print the config file path. |
qont config set |
Set a config value. |
qont config unset |
Remove a key (reverts to the built-in default). |
qont device¶
Browse, compare, and interrogate QPUs.
| Command | What it does |
|---|---|
qont device best |
Quick router: best devices for a qubit count and optional budget. |
qont device calib |
Per-qubit calibration table (T1/T2/readout) from bundled or @live data. |
qont device compare |
Two devices side by side. |
qont device drift |
Calibration weather report: today's live device vs the bundled snapshot. |
qont device list |
Every device in the pricing catalog. |
qont device queue |
Live queue depth for an IBM device (needs a saved account). |
qont device show |
Full catalog card for one device. |
qont device topology |
Sketch the qubit connectivity. |
qont env¶
Environment capture for reproducibility.
| Command | What it does |
|---|---|
qont env diff |
Compare the current environment against the lock file. |
qont env lock |
Write .qontinuum/env.lock.json capturing the current environment. |
qont env show |
Print the execution environment relevant to reproducibility. |
qont history¶
Query and mine the local run history.
| Command | What it does |
|---|---|
qont history bisect |
git-bisect for quantum: find the run where a check first went bad. |
qont history compare |
Statistic-level regression between two recorded runs. |
qont history export |
Export run summaries for spreadsheets or BI tools. |
qont history list |
Recent runs, newest last (run numbers are stable indices). |
qont history prune |
Trim history to the most recent N runs. |
qont history show |
Full detail of one run: every test and check with statistics. |
qont history stats |
Aggregate history: pass rate, flakiest test, shots and spend totals. |
qont lint¶
Static + probe-based checks for quantum test suites (rules Q001-Q009). Bare qont lint scans the current directory; qont lint check <path> scans elsewhere.
| Command | What it does |
|---|---|
qont lint |
Scan the current directory when no subcommand is given. |
qont lint check |
Scan a specific path. Exit 1 on findings. |
qont lint explain |
Explain one rule: what it catches and how to fix it. |
qont lint rules |
List every lint rule. |
qont noise¶
Noise models: inspect, sweep, and find headroom.
| Command | What it does |
|---|---|
qont noise headroom |
Load-testing for quantum: how much worse can the device get before this fails? |
qont noise inject |
Run one test under an explicit synthetic noise model. |
qont noise list |
Available noise sources. |
qont noise show |
The three-knob noise profile distilled from a device's calibration. |
qont noise sweep |
Run one test under progressively scaled device noise. |
qont pack¶
Create and verify reproducibility bundles.
| Command | What it does |
|---|---|
qont pack create |
Bundle the suite, baselines, seeded results, and environment lock. |
qont pack inspect |
Show a pack's manifest: environment, files, suite summary. |
qont pack verify |
Re-run the packed suite in THIS environment; do the results reproduce? |
qont providers¶
Pricing-catalog provenance.
| Command | What it does |
|---|---|
qont providers list |
Providers, device counts, verification date, and sources. |
qont providers verify |
Check that every catalog source URL is still reachable (network). |
qont plugin¶
Discover and diagnose provider, backend, and SDK plugins. See Extending Qontinuum.
| Command | What it does |
|---|---|
qont plugin list |
Every discovered plugin: name, kind, source, availability. |
qont plugin show |
Full detail for one plugin (disambiguate with --kind). |
qont plugin doctor |
Plugin health: what loaded, what needs an extra, what's broken. |
qont registry¶
Named circuit lineage (git for experiments).
| Command | What it does |
|---|---|
qont registry add |
Register a circuit (new version when the canonical hash changed). |
qont registry list |
Every registered circuit with its latest version. |
qont registry log |
Hash history, newest first. |
qont registry rm |
Remove a circuit and its whole lineage from the registry. |
qont registry show |
Latest version of one circuit. |
qont registry tag |
Tag a version (latest by default). |
qont report¶
Render suite results (from --from suite.json, or by running the suite).
| Command | What it does |
|---|---|
qont report badge |
SVG status badge from the suite result. |
qont report engineering |
Full engineering report: recommendation, plan, provider & cost analysis, history. |
qont report html |
Self-contained HTML dashboard from run history (alias of qont dashboard). |
qont report json |
The raw versioned suite-result JSON. |
qont report junit |
JUnit XML — Jenkins/GitLab/Buildkite render quantum checks natively. |
qont report md |
Markdown report (same renderer as the PR comment, without cost table). |
qont report pr |
The full sticky PR comment (results + hardware cost table), rendered locally. |
qont report summary |
One-line machine-greppable summary (for shell pipelines). |
qont schedule¶
Generate scheduled GitHub workflows (files only, never pushed).
| Command | What it does |
|---|---|
qont schedule nightly |
Emit a nightly (03:00 UTC) regression workflow. |
qont schedule weekly |
Emit a weekly (Monday 03:00 UTC) regression workflow. |
qont shots¶
Statistical shot-count planning.
| Command | What it does |
|---|---|
qont shots floor |
Smallest TVD threshold that this shot count can statistically resolve. |
qont shots for-tvd |
Minimum shots so a perfect device reliably passes the given threshold. |
qont shots plan |
Audit every discovered test: are its shots sound for its thresholds? |
qont shots power |
Smallest distribution drift a two-sample snapshot test can detect. |
qont snapshot¶
Manage golden-baseline snapshots.
| Command | What it does |
|---|---|
qont snapshot list |
List recorded golden baselines. |
qont snapshot prune |
Drop baselines whose tests no longer exist. |
qont snapshot rm |
Delete one baseline. |
qont snapshot show |
Full stored baseline for one test, including counts. |
qont snapshot update |
Re-record golden baselines for every snapshot-enabled test. |
qont snapshot verify |
Re-run snapshot tests and report drift without CI semantics. |
qont stats¶
Statistics on counts JSON files (flat dicts, runner output, or {'counts': ...}).
| Command | What it does |
|---|---|
qont stats chi2 |
Pearson goodness-of-fit test vs an expected distribution. |
qont stats compare |
Two-sample homogeneity test: were both files drawn from one distribution? |
qont stats describe |
Shots, support size, top outcomes, Shannon entropy. |
qont stats entropy |
Shannon entropy of the empirical distribution. |
qont stats fidelity |
Hellinger (classical) fidelity vs an expected distribution. |
qont stats tvd |
Total variation distance vs an expected distribution. |
qont telemetry¶
Opt-in, anonymous community intelligence (off by default). See the Quantum Intelligence Network and Privacy.
| Command | What it does |
|---|---|
qont telemetry status |
Show consent, endpoint, queued records, and community cache. |
qont telemetry enable |
Opt in to contributing anonymous execution metadata. |
qont telemetry disable |
Opt out; nothing further is captured or sent. |
qont telemetry preview |
Show exactly what would be contributed from your history. |
qont telemetry sync |
Flush queued records to the network (safe offline). |
qont telemetry community |
Show the locally cached community-intelligence snapshot. |
qont telemetry clear |
Delete all locally queued records without sending them. |
qont watch¶
Re-run on change / poll live calibration.
| Command | What it does |
|---|---|
qont watch device |
Poll live calibration and alert on drift (needs a saved IBM account). |
qont watch test |
Re-run the suite whenever a q_test file changes (Ctrl-C to stop). |