APIs and integrationsEdit on GitHubSource: docs/SURFACES.md

Where each capability is reachable

Every capability is available on the surfaces where it makes sense; where one cannot carry a capability, the cell says N/A and why. The same engine runs behind every surface: the Python, WebAssembly and MCP entry points for the 0.35 capabilities are thin wrappers over kshana::surface, which calls the code the command line runs.

A passage naming third-party AI clients is left out here; it is in docs/SURFACES.md on GitHub.

Capability CLI Rust Python WASM MCP Plugin JetBrains Notebook Studio
Vessel receiver-trust, batch log (0-100 score, reasons) receiver-trust receiver_trust::assess::assess_vessel_log receiver_trust, assess_vessel_log receiver_trust, assess_vessel_log assess_receiver_log, assess_vessel_log /kshana-assess-receiver, skill vessel-receiver-trust Assess Receiver Trust vessel-trust-and-training Maritime trust page
Receiver-trust on a bounded stream excerpt (scores, states, reasons; no gate) receiver-trust live receiver_trust::assess::assess_stream_excerpt receiver_trust_replay receiver_trust_replay assess_vessel_stream same Assess Receiver Stream (a stored log) vessel-trust-and-training Maritime trust page
Satellite geometry and deduction classes (0.36: geometry good/degraded/unknown with reasons and DOP, a label good/fair/poor always reported, the geometry monitor (nothing unless the geometry is poor, then up to 15 points; listed last in monitors_run when it decided, so the verdict's monitor count includes geometry), a class on every deduction, geometry_model, geometry_summary) receiver-trust, receiver-trust live (in the result JSON and every JSON line; the CSV, SVG, one-line verdict and $PKSHT carry a geometry deduction like any other and are otherwise unchanged) receiver_trust::geometry, on EpochTrust, Deduction and EpochReport receiver_trust, assess_vessel_log, receiver_trust_replay (in the returned dicts) the same three (in the returned JSON) assess_receiver_log, assess_vessel_log, assess_vessel_stream (in the reply; a vessel log's reply keeps geometry_model and geometry_summary); trust_telemetry renders it same, skills vessel-receiver-trust, trust-telemetry the actions above print the CLI output, which carries it vessel-trust-and-training (in the result it loads) pending: the Maritime trust page does not show it yet (fields listed in docs/MARITIME-TRUST.md, Studio view item 6)
receiver-trust live as a running process (stdin, followed file, TCP, UDP, --gate, --listen) receiver-trust live receiver_trust::live::LiveEngine N/A N/A N/A N/A N/A N/A Tools: Command line card (the command to run; a browser is not a process in the data path)
Evidence pack: create (signed, window of a log; 0.38: an optional lab-attestation, the operator's statement of a lab setup, recorded in the signed manifest, never verified) receiver-trust evidence (--lab-attestation <file.toml>) evidence::create_bundle, evidence::create_bundle_attested, surface::evidence_create, surface::evidence_create_attested evidence_create (lab_attestation) N/A (signing key) create_evidence_pack (lab_attestation_toml) /kshana-evidence-pack, skill evidence-pack Create Evidence Pack (no attestation field in the dialog in 0.38: pending; the command line takes it) vessel-trust-and-training N/A (a signing key does not belong in a page); Tools: Command line card
Jammertest as a referee (0.38): the published NPRA catalogue and transmission plan read as scheduled truth (import), a receiver log scored against it per test (score: detection latency, missed, false alarms, time to recover; MODELLED) jammertest import, jammertest score jammertest::plan, jammertest::score, surface::jammertest_import_json, surface::jammertest_score_json jammertest_import, jammertest_score jammertest_import, jammertest_score jammertest_import, jammertest_score /kshana-jammertest, skill jammertest N/A (needs the plan files and a log on disk; run it from a terminal) N/A (no notebook yet) pending: a Tools card for the command line (the Studio strip and views are WS-A's)
Signed Jammertest score pack (jammertest score --key: the receiver-trust evidence pack with the scorecard and the schedule inside config.json) jammertest score --key <keyfile> --out <dir> jammertest::score_pack N/A (signing key) N/A (signing key) N/A (signing key: use create_evidence_pack on a window, or the command line) N/A N/A N/A N/A (a signing key does not belong in a page); Tools: Command line card
Receiver self-report (0.38): UBX-NAV-STATUS spoofDetState and UBX-SEC-SIG read into self_report on each epoch, and the receiver_self_report comparison (both / receiver-only / Kshana-only / neither) in the result; raises no alarm and changes no score receiver-trust (in the result JSON; jammertest score) receiver_trust::ReceiverSelfReport, monitors::SelfReportEpoch, monitors::SelfReportSummary receiver_trust, assess_receiver_log (in the returned dicts) receiver_trust (in the returned JSON) assess_receiver_log (in the reply) /kshana-assess-receiver the actions above print the CLI output, which carries it N/A pending: the Trust page strip (WS-A; the field is epochs[].self_report)
Septentrio SBF log format (0.38, read only: PVTGeodetic, MeasEpoch, RFStatus) receiver-trust with [log] format = "sbf" receiver_trust::ingest::read_sbf, LogFormat::Sbf receiver_trust, assess_receiver_log (format = "sbf", log as base64) the same assess_receiver_log (base64) /kshana-assess-receiver the actions above N/A N/A (binary log upload not offered on the Trust page in 0.38)
receiver-trust live --serial <device> [--baud n]: a serial or USB NMEA feed, read only (opens the device for reading, never writes to the receiver; --configure is refused) receiver-trust live --serial N/A (a device and a process) N/A N/A N/A N/A N/A N/A Tools: Command line card (the command to run; a browser is not a process in the data path)
Evidence pack: create (signed, window of a log) receiver-trust evidence evidence::create_bundle, surface::evidence_create evidence_create N/A (signing key) create_evidence_pack /kshana-evidence-pack, skill evidence-pack Create Evidence Pack vessel-trust-and-training N/A (a signing key does not belong in a page); Tools: Command line card
Laboratory round-trip pack (scenario, test conditions, receiver log and score linked by hash) evidence round-trip evidence::build_round_trip_pack, surface::evidence_round_trip evidence_round_trip N/A (signing key) create_round_trip_pack /kshana-evidence-pack N/A (JetBrains actions are bound to one selected file; the pack needs a scenario, conditions, a log and a key: run it from the IDE terminal) N/A (a signing key does not belong in a page); Tools: Command line card
Evidence pack: verify evidence verify evidence::verify_bundle evidence_verify evidence_verify verify_evidence_pack same Verify Evidence Pack (pinned key only) vessel-trust-and-training Tools: Evidence packs (verify, attach timestamp)
Trust telemetry (Prometheus, syslog CEF/LEEF, OTLP; 0.38: format = "mqtt", Home Assistant discovery and state lines; --mqtt-tcp publishes from the CLI) trust-telemetry telemetry trust_telemetry trust_telemetry (text out; no socket) trust_telemetry /kshana-trust-telemetry, skill trust-telemetry Render Trust Telemetry (a stored result) vessel-trust-and-training Tools: Telemetry card (Command line tab; formats only, no sockets)
Machine and sensor inputs (0.38): receiver-trust formats rtklib-pos, novatel, gnss-sdr, and [log.imu] (csv, euroc) receiver-trust (scenario [log] format) receiver_trust::machines receiver_trust, assess_receiver_log (inline) receiver_trust (inline) assess_receiver_log (inline text/base64; a path is refused) /kshana-assess-receiver Assess Receiver Trust vessel-trust-and-training N/A (Studio's Trust page reads the formats it lists; new formats follow with its next port)
IMU log to hybrid/fusion sensor parameters ([accel_classical] table from a stationary log, Allan fit) imu-fit surface::imu_fit, receiver_trust::machines::imu_fit imu_fit imu_fit fit_imu_log /kshana-fit-imu N/A (no IDE action yet) N/A N/A (Studio's scenario editor takes the pasted table)
ROS 2 node, GNSS-SDR UDP relay, trust-telemetry --mqtt-tcp (integrations/ros2, integrations/gnss-sdr) processes N/A N/A N/A N/A N/A N/A N/A N/A (long-running processes and sockets; text forms: trust_telemetry with mqtt)
Galileo OSNMA verification of I/NAV pages or a UBX stream (docs/OSNMA.md) osnma verify osnma::cli::verify_report, surface::osnma_verify osnma_verify osnma_verify N/A (a long raw receiver stream suits the CLI and Python; an agent tool would add a large-input surface for no benefit) N/A (plugin commands run over the MCP server, which has no OSNMA tool for the reason in the MCP cell; run kshana osnma verify) N/A (the result means nothing without a trusted Merkle root or public key obtained from the ICD by another route and, for freshness, an independent reference time; a right-click action has nowhere to take them, and an unpinned default would read as verification; run kshana osnma verify) N/A (a stream in, a JSON report out; no notebook) Tools: OSNMA
Training NMEA with an instructor log nmea-scenario nmea_synth::generate_from_toml nmea_training nmea_training generate_training_nmea /kshana-training-scenario, skill training-nmea-scenario Generate Training NMEA vessel-trust-and-training Training debrief page
Training NMEA streamed to a TCP or UDP address nmea-scenario --tcp/--udp nmea_synth::stream N/A N/A N/A N/A N/A N/A Tools: Command line card (streams to an address)
Interference map (ADS-B, AIS) interference-map adsb|ais interference_map::api interference_map interference_map build_interference_map /kshana-interference-map, skill interference-map-route-exposure Build Interference Map interference-map-route-exposure Interference map page
Emitter region (ADS-B, AIS; MODELLED) interference-map locate interference_map::api::emitter_regions_from_adsb_csv, emitter_regions_from_readsb_traces, emitter_regions_from_ais_csv interference_emitter_regions interference_emitter_regions N/A (not yet carried; the Rust, Python and WASM surfaces cover it) N/A N/A N/A (publishing emitter regions in the public observatory is a separate owner decision and is off)
Route exposure route-exposure interference_map::api::route_exposure route_exposure route_exposure route_exposure same Route Exposure interference-map-route-exposure Interference map page
Signal K plugin and OpenCPN gate feed (docs/MARINE-INTEGRATIONS.md) integrations/, receiver-trust live --gate --listen N/A (processes) N/A N/A N/A how-to only: /kshana-marine-integrations, skill marine-integrations N/A N/A Tools: Command line card (how-to; a deployment)
Test-bench export (motion CSV, NMEA, waypoints, events for a laboratory simulator) bench-export interop::testbench::export_with_notes, surface::bench_export bench_export bench_export export_test_bench /kshana-bench-export, skill test-bench-export Export Test-Bench Files compliance-and-test-bench Tools: Test-bench export
Compliance mapping from runs (which public-framework rows the runs support evidence for) compliance-report compliance::assess_texts, surface::compliance_report compliance_report compliance_report compliance_report /kshana-compliance-report, skill compliance-mapping Build Compliance Report compliance-and-test-bench Tools: Compliance mapping
Compliance mapping tables and sources compliance-report --mapping, --sources compliance::mapping, surface::compliance_mapping compliance_mapping compliance_mapping compliance_mapping same Show Compliance Mapping N/A Tools: Compliance mapping
Standards packs (scored scenario sets with bars stated before the run: DHS RPCF test profiles, IMO MSC.1/Circ.1575, ICAO Doc 9849 Appendix F) standards-pack standards_pack, surface::standards_pack_list, surface::standards_pack_run standards_pack_list, standards_pack_run standards_pack_list, standards_pack_run list_standards_packs, run_standards_pack /kshana-standards-pack N/A (JetBrains actions are bound to a selected file; packs are embedded and need none: run kshana standards-pack from the IDE terminal) N/A (Studio Tools cards for packs belong to the Studio workstream; the CLI or any surface output feeds the Compliance mapping card)
Data-partner submission check (format and metadata of a measured-data submission, kshana.partner-submission/1; scores nothing) partner check partner::check_file, partner::check N/A (the check reads local files and their SHA-256; the Rust API is the library surface and the CLI is the tool) N/A (no file system in the browser build) N/A (MCP refuses file sources by design (inline text only)) N/A (slash commands call MCP tools, and this check needs files on disk) N/A (JetBrains actions are bound to one selected file; a submission is a folder) N/A (no notebook yet) N/A (Studio has no file-system access; run the command)
LunaNet Augmented Forward Signal: codes, SB1/CRC/interleaver, and a check against LSIS-AFS V1.0 (docs/LUNANET-AFS.md) lunanet-afs-check; signals lunanet-afs-i, lunanet-afs-q in iq scene/acquire/track/sweep iq::signals::lunanet, iq::lunanet_afs, surface::lunanet_afs_check_spec, lunanet_afs_check_symbols, lunanet_afs_generate lunanet_afs_check_spec, lunanet_afs_check_symbols, lunanet_afs_generate; iq_scene etc. with the two signal names lunanet_afs_check_spec, lunanet_afs_check_symbols, lunanet_afs_generate; iq_scene etc. through the short in-memory runs lunanet_afs_check /kshana-lunanet-afs-check N/A (JetBrains actions are bound to a selected file; run kshana lunanet-afs-check from the IDE terminal) N/A (a Studio card is a Studio-workstream item; the CLI and every other surface carry it)
Sky geometry under a skyline: visible and masked satellites, DOP, ENU covariance, error ellipse (docs/SKYLINE.md) sky geometry::sky_geometry, surface::sky_geometry sky_geometry sky_geometry sky_geometry /kshana-run (points to the tool) N/A (an interactive skyplot is a browser view; kshana sky covers the numbers) notebooks/geometry.ipynb — (the Studio Geometry view is added separately in 0.36)
Sky Monte Carlo cloud (seeded least-squares errors, RMS, 95 % containment) sky --monte-carlo <n> --seed <s> geometry::sky_monte_carlo sky_monte_carlo sky_monte_carlo N/A (bulk samples an agent should summarise, not carry) N/A (no MCP tool) N/A (as above) N/A — (as above)
Skyline presets and profiles (docs/SKYLINE.md) skyline --preset <name>, --presets, --profile <file> geometry::skyline_presets, geometry::skyline_profile skyline_presets, skyline_profile skyline_presets, skyline_profile sky_geometry (skyline_preset) same N/A (as above) N/A — (as above)
Skyplot as SVG sky --svg geometry::skyplot_svg skyplot_svg skyplot_svg N/A (an agent reads the numbers from sky_geometry) N/A N/A (as above) N/A — (as above)
Skyline section in a scenario ([skyline]: orbit and integrity with user_site, gnss-sim, jamming, leo-pvt, leo-pass, nmea-scenario) the scenario file skyline::SkylineCfg run, nmea_training run, nmea_training run_scenario, generate_training_nmea /kshana-run Run Kshana Scenario N/A —
Optical timing link through the atmosphere: per-site availability, site diversity, enhancement techniques, outage-to-holdover coupling (scenario kind optical-atmosphere; docs/OPTICAL-ATMOSPHERE.md; MODELLED, ILLUSTRATIVE sites, advisory) kshana scenarios/optical-atmosphere-network.toml, scenarios/optical-atmosphere-techniques.toml optical_atmosphere, optical_weather, optical_sitediv, optical_techniques, optical_outage_holdover run, run_typed run run_scenario /kshana-run Run Kshana Scenario optical-atmosphere pending: no Studio view yet; the scenario runs in the Studio scenario runner and the Tools Command line card
Holdover test scored from an operator-supplied time-error record: baseline, GNSS removed by the operator, holdover, recovery, measured against ITU-T masks (scenario kind holdover-test; docs/TIMING-HIL.md; MODELLED, advisory; Kshana transmits and drives nothing) kshana scenarios/holdover-test-example.toml, scenarios/holdover-test-ptp4l.toml holdover_test, telecom_timing run, run_typed run (inline samples or clock_log.text; the file keys are refused) run_scenario (inline only) /kshana-run Run Kshana Scenario N/A (no notebook in 0.38; the example scenarios run in any notebook through run) pending: no Studio view yet; the scenario runs in the Studio scenario runner and the Tools Command line card
Navigation and timing error budget of a chain of mission phases from launch to entry: eight phase kinds, handover rules, frame and timescale conversions, a ground-network overlay, margin against stated requirements and the weakest link (scenario kind mission-thread; docs/MISSION-THREAD.md; MODELLED, ILLUSTRATIVE inputs, advisory; public sources only, no endorsement) kshana scenarios/mission-thread-artemis-class.toml (from a checkout: it reads a committed Moon kernel cut) mission_thread run, run_typed run (moon_kernel_path is a file key and is refused; without a kernel the Moon is the analytic series and the conversion row says so) run_scenario (inline only; the kernel key is refused) /kshana-run Run Kshana Scenario N/A (no notebook; the example runs in any notebook through run from a checkout) pending: no Studio view yet; the scenario runs in the Studio scenario runner and the Tools Command line card
GNSS beyond the constellation: availability of GPS, GLONASS, Galileo, BDS, QZSS and NavIC signals for a user on a shell or on a LEO to lunar-transfer trajectory (booklet, parametric or measured transmit pattern, limb-ray medium, weak-signal receiver design), an onboard GNSS EKF for high-orbit and lunar-transfer craft, and formation relative navigation (scenario kind gnss-ssv; docs/SSV.md; MODELLED, advisory; compared with the ICG SSV booklet under a pre-registration, xval/gnss-ssv/RESULTS.md) kshana scenarios/gnss-ssv-shell.toml, gnss-ssv-heo.toml, gnss-ssv-lunar-weak.toml, gnss-ssv-nav-lunar.toml, gnss-ssv-formation-heo.toml ssv run, run_typed run (inline; transmit.pattern_text carries a measured pattern inline) run_scenario (inline only; transmit.pattern_csv is refused) /kshana-run Run Kshana Scenario N/A (no notebook in 0.38; the scenarios run in any notebook through run) pending: no Studio view yet; the scenario runs in the Studio scenario runner and the Tools Command line card
Lunar orbit and surface as navigation and geodetic infrastructure: orbit-determination consistency, carrier-phase ambiguity resolution, frame co-estimation, space-tie and Love-number sensitivity (scenario kind lunar-infrastructure, study = ...; docs/LUNAR-INFRASTRUCTURE.md; MODELLED, ILLUSTRATIVE inputs, advisory) kshana scenarios/lunar-infrastructure-od-consistency.toml (and the five sibling examples) lunar_infra run, run_typed run run_scenario /kshana-run Run Kshana Scenario N/A (no notebook; the examples run in any notebook through run) pending: no Studio view yet; the scenarios run in the Studio scenario runner and the Tools Command line card
Measured clock logs read to a phase series: linuxptp ptp4l, phc2sys, ts2phc, chrony tracking and statistics, beside RINEX clock, TimeLab and TICC ([clock_log] / [series.clock_log]; docs/CLOCK-LOG-FORMATS.md; MODELLED, no oracle compares a parse with another reader) a scenario with [clock_log] clock_log::read_input run run (text inline only; no file reads in a browser) run_scenario (text inline only; path is refused) /kshana-run Run Kshana Scenario N/A (no notebook; a log is a file the user supplies) pending: no Studio view; the Studio scenario runner takes the scenario text
Optical timing link through the atmosphere: per-site availability, site diversity, enhancement techniques, outage-to-holdover coupling (scenario kind optical-atmosphere; docs/OPTICAL-ATMOSPHERE.md; MODELLED, ILLUSTRATIVE sites, advisory) kshana scenarios/optical-atmosphere-network.toml, scenarios/optical-atmosphere-techniques.toml optical_atmosphere, optical_weather, optical_sitediv, optical_techniques, optical_outage_holdover run, run_typed run run_scenario /kshana-run Run Kshana Scenario optical-atmosphere the Studio's Optical PNT view (/studio/optical/, 0.38) draws the two bundled scenarios' recorded results, including the seeded time series; the scenario also runs in the Studio scenario runner and the Tools Command line card
Land-polygon download (interference-map fetch-land) --allow-network CLI code N/A N/A N/A N/A N/A N/A Tools: Command line card (needs --allow-network)
Benchmark (kshana-bench v1: score a submission against a truth file, leaderboard page; docs/BENCH.md, MODELLED, advisory) bench score, bench leaderboard bench::score, bench::site::leaderboard_html, surface::bench_score, surface::bench_leaderboard bench_score, bench_leaderboard bench_score, bench_leaderboard score_benchmark, render_leaderboard N/A (the two MCP tools carry it; a slash command over two JSON files adds nothing) N/A (not an IDE action) N/A (education notebooks are a separate workstream) N/A (the leaderboard is the static page bench/site/, ported to the site as it is; Studio computes nothing)
Benchmark data and holdout tools (bench truth|cases|baseline|seal|reveal-check|manifest|verify-manifest|validate) the same bench::tasks, bench::seal N/A N/A N/A N/A N/A N/A N/A (the CLI reads dataset directories and clock files and, for seal, a secret seed file; a seed must not pass through an agent or a page)
Observatory daily product and sensitivity study (interference-map daily|sensitivity; docs/observatory/METHOD.md) interference-map daily, interference-map sensitivity interference_map::observatory N/A N/A N/A N/A N/A N/A N/A (the commands read an archive directory of hundreds of thousands of files, write four files including a PNG, and enforce the 24 h release delay against the clock of the machine that publishes; the CSV map is on every surface in the row above)

Studio#

Kshana Studio is the web app built on the WebAssembly build, so a Studio view exists where the WASM column has an entry point, and where it does not the cell points to a Command line card (the command to run, with its arguments), never to a result Studio computed. Pages (kshana.dev/studio/, client-side): Maritime trust (trust/), Training debrief (training/), Interference map (interference/), Geometry (geometry/: a representative skyplot, skyline, DOP, error ellipse and Monte Carlo cloud from the engine; 0.36), Lunar station (lunar/: the modelled coverage of a lunar reference station from the lunar-station-coverage kind; 0.36) and the Tools page, whose sections are OSNMA (#osnma), Evidence packs (#evidence), Test-bench export (#bench), Compliance mapping (#compliance), the Command line tab (telemetry card, IQ, marine integrations: commands to run) and Use from your AI assistant (#ai). A — in the Studio column of the subcommand table is a command that is not one of the 0.35 capabilities Studio presents.

Why the N/A cells are N/A#

  • Signal K and OpenCPN. They are consumers of the running live process and of its TCP gate, so they are command-line deployments; the plugin carries a how-to command and a skill that point to docs/MARINE-INTEGRATIONS.md and call no tool.

  • A running process. receiver-trust live holds a stream open, can serve the gated stream to any number of TCP clients and runs until it is stopped. A Python call, a browser call, an MCP tool call and an IDE action each return when they finish and do not hold a socket open for others. Python, WASM and MCP carry the bounded form instead: an excerpt scored by the same engine, in the live schema (JSON lines 1.3: position, an advisory statement and the satellite geometry on every epoch, a class on every deduction), with no socket and no gate (the gate marks a fix invalid on a live stream, which only a process in the data path can do). Two behaviours belong to the process and are not in the excerpt tools: a stream replayed faster than real time through receiver-trust live without --replay is scored untrusted (its time disagrees with the host clock; an excerpt is read without a host clock), and --udp <port> and --listen tcp:<port> bind the loopback address unless an address is given.

  • Trust telemetry. WASM carries trust_telemetry(input, format, host, input_kind, options_json): the same encoders, returning Prometheus text, CEF or LEEF syslog lines, or the OTLP JSON payload for a stream or a batch result, with no clock (syslog lines carry the RFC 5424 nil timestamp) and nothing sent. Python (trust_telemetry) and MCP (trust_telemetry, with the plugin command /kshana-trust-telemetry) carry it too. The rest of trust-telemetry serves Prometheus metrics, sends syslog and exports OpenTelemetry from a running stream, so it is a process by nature. How-to: pipe the live command into it, or replay a result file (kshana trust-telemetry --result session.result.json --print-metrics); see TRUST-TELEMETRY.md. An agent or a notebook that wants the per-epoch scores calls the batch or excerpt functions instead.

  • Creating a pack in the browser. Signing needs a private key, which should not pass through a web page; the browser build verifies packs (evidence_verify) and the other surfaces create them. The MCP tool never returns or logs the signing seed, and without one it signs with a one-time key (integrity, not identity); for a key that matters, use the CLI. The same holds for streaming training NMEA to an address: Python, WASM and MCP return the generated text, and the CLI sends it.

  • Study suites. --study <suite.toml> runs a manifest that names sibling scenario files on disk. The in-memory surfaces take scenario text, not a folder of files, so they run each member as a scenario instead (run, run_scenario, run_all).

  • GNSS IQ in the browser, and recordings that are rewritten. A recording is gigabytes of samples, which a page cannot hold and the MCP protocol must not carry, so WASM has no IQ functions and the MCP tools read and write one work directory only (KSHANA_MCP_IQ_DIR, with a per-call sample budget). iq extract, convert and decimate exist to rewrite recordings on disk and stay command-line only. iq labfit and iq conditions carry text, so MCP takes them inline (a log that names a file is refused).

  • The land download. It is the only command that opens a network connection, it needs the system curl and an explicit --allow-network, and it checks a pinned SHA-256. The browser has no curl, and an AI agent should not fetch data on a user's behalf; the other surfaces take the land file as text instead (land_geojson).

  • Trace archives and directories. interference-map adsb on a readsb trace directory reads files from disk and is command-line only; the in-memory surfaces take CSV text.

Every command-line subcommand, and who covers it#

One row per subcommand or mode of kshana --help. yes names the entry point; no is a gap with its reason; N/A is a surface that cannot carry it for the reason given above or in the last column. JetBrains actions run the command line, so an N/A there states why no IDE action exists.

Subcommand Python WASM MCP Plugin JetBrains Notebook Studio Note
kshana <scenario.toml> (run, chart, table) run, run_full, run_typed run, chart_svg, summary, table_csv, run_all run_scenario, report_scenario, export_table_csv /kshana-run Run Kshana Scenario quantum-vs-classical-gdop —
--validate <scenario.toml> validate_toml error_kind (a run throws on an invalid scenario) validate_scenario /kshana-run Validate Scenario —
--study <suite.toml> N/A N/A N/A N/A N/A (names sibling files on disk) — A suite names sibling scenario files on disk (N/A above); run each member as a scenario
--export-sp3/-omm/-oem export_sp3, export_omm, export_oem export_sp3, export_omm, export_oem export_sp3, export_omm, export_oem /kshana-run Export Scenario —
--export <czml,kml,geojson,stk,sigmf> export_scenario, export_formats export_scenario, export_formats export_interop, list_export_formats /kshana-run Export Scenario — Files come back as text (utf-8) or base64 (the SigMF samples)
--import-route import_route import_route import_route /kshana-import-route Import Route —
--animate <svg,html,frames> animate_scenario animate_scenario animate_scenario /kshana-run Animate Scenario — At most 120 frames per call; longer sequences use the CLI
sky <request> (or flags), --monte-carlo, --svg sky_geometry, sky_monte_carlo, skyplot_svg sky_geometry, sky_monte_carlo, skyplot_svg sky_geometry (no Monte Carlo, no SVG) /kshana-run N/A (see the capability row) — MODELLED geometry; JSON or SVG out
skyline --preset, --presets, --profile skyline_presets, skyline_profile skyline_presets, skyline_profile sky_geometry (skyline_preset) /kshana-run N/A (see the capability row) — REPRESENTATIVE presets with provenance
kinds [--json] scenario_kinds, list_kinds list_kinds list_scenario_kinds /kshana-run List Scenario Kinds —
example [<name>] list_examples, get_example list_examples, get_example list_example_scenarios, get_example_scenario /kshana-run Insert Example Scenario — The python and wasm features compile the example text in (about 0.7 MB)
receiver-trust <scenario.toml> (batch) receiver_trust, assess_vessel_log receiver_trust, assess_vessel_log assess_receiver_log, assess_vessel_log /kshana-assess-receiver Assess Receiver Trust vessel-trust-and-training Maritime trust page
receiver-trust live excerpt: receiver_trust_replay excerpt: receiver_trust_replay excerpt: assess_vessel_stream excerpt Assess Receiver Stream (a stored log, --file) Maritime trust page (excerpt); Command line card for the process A running process; the excerpt has no socket and no gate (N/A above)
trust-telemetry trust_telemetry trust_telemetry trust_telemetry /kshana-trust-telemetry Render Trust Telemetry (a stored result) vessel-trust-and-training Tools: Telemetry card (Command line tab; formats only, no sockets) Text only on every surface, no socket; the running exporters are a process (see above)
imu-fit <imu.csv> imu_fit imu_fit fit_imu_log /kshana-fit-imu N/A — Text in, JSON out; MODELLED
osnma verify <pages|vectors.csv|stream.ubx> osnma_verify osnma_verify N/A (see above) N/A (see above) N/A (see above) Tools: OSNMA Same JSON as --json, with the advisory in its advisory field; the file read becomes bytes in memory. Authentication status only: not a navigation integrity service, and no anti-spoofing performance claim
receiver-trust evidence (create a pack) evidence_create N/A create_evidence_pack /kshana-evidence-pack Create Evidence Pack vessel-trust-and-training Tools: Command line card (no signing key in a page) No signing key in a browser
jammertest import, jammertest score jammertest_import, jammertest_score jammertest_import, jammertest_score jammertest_import, jammertest_score /kshana-jammertest N/A (terminal; needs files) pending: Tools: Command line card Plan and catalogue are given as text; the signed pack (--key) is command line only (signing key)
evidence round-trip evidence_round_trip N/A create_round_trip_pack /kshana-evidence-pack N/A (as above) Tools: Command line card (no signing key in a page) No signing key in a browser; no signal is written or replayed
evidence verify evidence_verify evidence_verify verify_evidence_pack /kshana-evidence-pack Verify Evidence Pack (pinned key only) vessel-trust-and-training Tools: Evidence packs (verify, attach timestamp)
evidence keygen evidence_keygen N/A N/A N/A N/A (a private key is never made or read by the plugin) Tools: Command line card Python returns the key in memory with a warning; a private key should not pass through a conversation or a page
evidence attach-timestamp evidence_attach_timestamp evidence_attach_timestamp attach_evidence_timestamp /kshana-evidence-pack Attach Timestamp Tools: Evidence packs (verify, attach timestamp) Token as bytes or base64, updated pack returned in memory; the authority's signature is not checked (openssl ts -verify)
nmea-scenario <scenario.toml> nmea_training nmea_training generate_training_nmea /kshana-training-scenario Generate Training NMEA vessel-trust-and-training Training debrief page
nmea-scenario --tcp/--udp N/A N/A N/A N/A N/A (see Note) Tools: Command line card Sends to an address (N/A above)
interference-map adsb|ais (CSV) interference_map interference_map build_interference_map /kshana-interference-map Build Interference Map interference-map-route-exposure Interference map page
interference-map adsb (trace directory) N/A N/A N/A N/A N/A (see Note) Tools: Command line card Reads a directory (N/A above)
interference-map locate (CSV) interference_emitter_regions interference_emitter_regions N/A (not yet carried) N/A N/A — Estimated emitter regions, MODELLED; trace directories and --tdoa are command line only
interference-map fetch-land N/A N/A N/A N/A N/A (see Note) Tools: Command line card Network access (N/A above)
interference-map daily, sensitivity N/A N/A N/A N/A N/A (see Note) — Reads an archive directory and writes files; the release delay is enforced on the publishing machine (N/A above)
bench score bench_score bench_score score_benchmark N/A (the MCP tool carries it) N/A — Text in, text out; nothing read or written
bench leaderboard bench_leaderboard bench_leaderboard render_leaderboard N/A (the MCP tool carries it) N/A the static page bench/site/ HTML text
bench truth|cases|baseline|seal|reveal-check|manifest|verify-manifest|validate N/A N/A N/A N/A N/A — Dataset files and a secret seed (N/A above)
route-exposure route_exposure route_exposure route_exposure /kshana-interference-map Route Exposure interference-map-route-exposure Interference map page
bench-export bench_export bench_export export_test_bench /kshana-bench-export Export Test-Bench Files Tools: Test-bench export Returns text, writes nothing; no signal
compliance-report compliance_report compliance_report compliance_report /kshana-compliance-report Build Compliance Report Tools: Compliance mapping
compliance-report --mapping, --sources compliance_mapping compliance_mapping compliance_mapping /kshana-compliance-report Show Compliance Mapping —
iq scene iq_scene, iq_scene_broadcast N/A iq_scene /kshana-iq N/A (recordings are gigabytes of files; see Note) — Samples are files. --format sdrx[:ci16_le] (or an out path ending .sdrx) writes ION GNSS SDR metadata beside the samples: the CLI and MCP (format: "sdrx") write it; Python returns the samples in memory and writes no file, so the metadata has no Python form (N/A)
standards-pack list, run, rows, tests standards_pack_list, standards_pack_run standards_pack_list, standards_pack_run list_standards_packs, run_standards_pack /kshana-standards-pack N/A (as above) Tools: Command line card Scenarios are embedded; no file system; rows and tests print the committed doc tables and are CLI-only
partner check <submission.toml> N/A N/A N/A N/A N/A — Reads a submission file and its data file on disk and verifies the SHA-256; none of those surfaces has the folder. Format and metadata only, nothing scored; see docs/PARTNER-DATA.md
lunanet-afs-check spec, symbols, generate lunanet_afs_check_spec, lunanet_afs_check_symbols, lunanet_afs_generate lunanet_afs_check_spec, lunanet_afs_check_symbols, lunanet_afs_generate lunanet_afs_check /kshana-lunanet-afs-check N/A (as above) — Symbol streams are text of 0 and 1, at most 4 MiB (MCP) or 100 frames generated per call; LDPC and CRC-24Q of subframes 2 to 4 are not verified; no signal is written
iq scene iq_scene, iq_scene_broadcast N/A iq_scene /kshana-iq N/A (recordings are gigabytes of files; see Note) — Samples are files
iq acquire iq_acquire, iq_acq_surface N/A iq_acquire /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq track iq_track, iq_loop_designs N/A iq_track /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq sweep iq_sweep N/A iq_sweep /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq labfit iq_labfit N/A iq_labfit (logs inline, no files) /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq frontend iq_frontend N/A iq_frontend /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq monitor iq_monitor N/A iq_monitor /kshana-iq N/A (recordings are gigabytes of files; see Note) — --follow [--idle-timeout <s>] reads a SigMF recording that is still being written, receive only (iq::io::follow::open_sigmf_follow in Rust): N/A for Python and MCP, because a call that waits for a file to grow does not return when the work is done, the reason receiver-trust live is also command-line only; the recording at its current length runs through iq_monitor
iq campaign, iq campaign report iq_campaign, iq_campaign_report N/A iq_campaign, iq_campaign_status /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq conditions iq_test_conditions N/A iq_test_conditions (text, inline) /kshana-iq N/A (recordings are gigabytes of files; see Note) —
iq inventory, info iq_inventory, iq_info N/A iq_info /kshana-iq N/A (recordings are gigabytes of files; see Note) — Inventory of a folder is Python and CLI; the MCP server reads one work directory
iq extract, convert, decimate N/A N/A N/A N/A N/A (see Note) — They rewrite recordings on disk (N/A above)
the signal list (iq --help) iq_signals N/A iq_signals /kshana-iq N/A (a list; run kshana iq --help) —

Utility functions#

Small entry points that are not a capability row, and where each lives:

Function Surface What it is
version Python (kshana.version(), kshana.__version__), WASM (version), CLI (--version), MCP (the server info) The engine version
iq_read_epochs Python Reads a kshana.track-epoch/1 binary file that kshana iq track --epochs wrote. N/A elsewhere: it reads a file from disk (WASM has no file system, MCP reads one work directory only through the iq_* tools)
encode_permalink, decode_permalink WASM The ?s= share-link codec of the browser playground (scenario TOML to a URL-safe token and back). N/A elsewhere: a link to a web page has no meaning for a command line, a Python call or an agent

Limits that differ by surface#

Surface Input limit Files
MCP 4 MiB per text input (2 MiB and 20,000 epochs for a stream excerpt); replies carry at most 2000 NMEA lines, 200 non-nominal epoch lines and 4 MiB of GeoJSON or pack files inline content only: a scenario field that names a file is refused; the iq_* tools use one opt-in work directory, with every path confined to it
MCP over HTTP (--http, self-hosted) as MCP, with a 5 MiB request body, 120 s per request and 8 requests at once by default none: the iq_* file tools are off; every other tool takes inline content
Python, Rust 64 MiB per text input receiver_trust may name a path the process can read
WASM 64 MiB per text input, inside the browser's own memory none: the browser has no file system, nothing is uploaded

Evidence tiers and caveats on every surface#

  • An evidence pack is a technical record of what the engine computed from a log under stated settings, with every hash and a signature over them. It is not a legal opinion, a finding of fact about any event or an approval. A pass of verify says the record is unchanged; with no trusted public key it proves only integrity against the key the pack names itself.
  • The receiver-trust monitors and the trust score are MODELLED: thresholds are stated inputs and nothing asserts how they perform against real interference. The output is advisory; Kshana is not type-approved navigation equipment (IEC 61108, IEC 61162) and the operator remains responsible. A high score is not proof of a genuine fix.
  • The interference map is MODELLED (pre-registered thresholds, not validated against a ground-truth measurement). A degraded cell does not identify interference as the cause; a cell not observed is not evidence of a clear route; it describes past reports and is not a forecast. It is aggregate only: no identifier is returned.
  • Training NMEA is text only: no RF, IQ or waveform is synthesised, nothing is transmitted, and it is never for a vessel's live navigation systems.

The evidence label, and the outputs that carry none#

Every figure and table a run writes carries its evidence tier (VALIDATED, MODELLED or PARTNER, the weakest of the ledger rows behind it) and the ids of those rows, on every surface that runs a scenario: the result JSON (evidence key), the chart SVG (metadata element and a line under the plot) and the CSV table (# evidence: comment lines). The same label is on the receiver-trust outputs (CLI, MCP, Python), jammertest score --json, iq acquire, iq sweep, iq track --summary, iq monitor --json, the interference-map route report, and inside the config.json of every evidence pack (signed; kshana evidence verify reports it). Row ids are stable (VM- and eight hex digits, the ID column of docs/VERIFICATION-MATRIX.md).

These outputs carry no label, each for a stated reason (the list is UNLABELLED_WRITERS in tests/evidence_labels_doc_sync.rs, which fails on a writer that is neither labelled nor listed):

Output Why it carries no label
kshana-bench result files a fixed schema pinned by the holdout commitments; each carries its own label (MODELLED or measured)
interference observatory daily products (GeoJSON, PNG, incident JSON) a frozen, published schema; the method note's caveats travel with them
IQ campaign cell, scorecard and report files digest-pinned; each states the campaign scorecard it restates
IQ samples, SigMF and SDR metadata, scene truth, track epoch and event streams, live session logs format-bound files and record streams read by other tools
SP3, OMM, OEM, CZML, KML, GeoJSON, STK, test-bench motion and NMEA exports; the NMEA training stream format-bound files read by third-party tools
telemetry wire formats (Prometheus, OTLP, CEF, LEEF) defined by their wire formats
generated documents (the ledger itself, the committed study and data tables) not the output of a run