CLI reference

Generated

This page is generated from the code when the documentation is built (docs/hooks/reference.py). Do not edit it; change the code instead.

Every command of cheme, with its arguments and options, as cheme <command> --help prints them. The guides explain how the commands fit together.

command	what it does
cheme init	Create a plant repository from the template (folders, schemas, skills, CI).
cheme validate	Validate every P&ID and list, and the lists against the P&IDs.
cheme fmt	Write P&IDs in canonical form: readable ids on every object, DEXPI 2.0.1, stable order.
cheme render	Render P&IDs to SVG or PNG: build/<sheet>(@<rev>)(_crop-X0-Y0-X1-Y1)(_inline).svg|png.
cheme diff	Compare P&IDs between two git revisions; write an HTML report with changes highlighted.
cheme show	Where an object is: its block, owner, drawing, references and what references it.
cheme check	Run the engineering-step gates in checks/checks.yaml (--write signs off a stage).
cheme status	Where the repository stands: sheets, findings by sheet, gates, sign-offs, review findings.
cheme impact	Artifacts whose sources changed since they were reviewed or generated: review needed.
cheme maturity	Which calculation maturity level the evidence supports, requirement by requirement.
cheme issues	The plant's issues (plant.yaml issues:) as engineering items: actions, holds, technical queries and design changes,...
cheme cause-effect	Write the cause-and-effect matrix of control/cause-effect.yaml (HTML and CSV).
cheme serve	Serve the repository read-only as a JSON API (and the viewer, with --ui).
cheme export-static	Write the viewer's API as static files: a read-only snapshot the viewer opens without a server (manifest.json + data/).
cheme git-setup	Show P&IDs as review text in git diffs, and check them before each commit.
cheme schema	Print the JSON Schema of a YAML file kind, or list the kinds.
cheme migrate	Bring the YAML inputs to the schema versions this cheme-as-code reads: a file without schema: gets `schema:...
cheme copy	Duplicate an object with its parts and drawing, renaming every id (e.g.
cheme fragment	Print a ready-to-paste XML fragment (conceptual part and drawing part).
cheme run	Run a Python script in cheme-as-code's interpreter, where pyDEXPI and chemeascode import.
cheme regen	Rebuild generated sheets: frame -> cheme run SCRIPT FILE -> fmt.
cheme move	Move an object's drawing: symbol, node positions, labels; attached lines follow.
cheme rename	Rename an id and every reference to it, or a tag and the labels that show it.
cheme set	Set data values of an object (typed as DEXPI defines them) and the labels showing them.
cheme sheet	Start a P&ID sheet with a real title block, or change its title block.
cheme process	Process design: BFD and PFD generated from process/materials.yaml and process/flowsheet.yaml.
cheme layout	3D layout: the plant model generated from layout/*.yaml, the lists and the P&IDs.
cheme assess	Assessment: the material take-off of the 3D model, the estimates, the LCA and the TEA.
cheme datasheets	Equipment datasheets: drafts, calculated sizes and dimensions, outline drawings.
cheme skills	The bundled skills installed in .claude/skills/: status and safe updates.
cheme chemistry	Chemistry: component properties from databases and correlations.
cheme control	Control: the loop register, operating sequences, narrative and control philosophy.
cheme safety	Safety: HAZOP studies and their risk matrix.
cheme experiments	Experiments: kinetic data, rate laws fitted to them, data sampled from a mechanism, measured electrochemical series.
cheme data	Optional datasets: fetch them into the user data directory.
cheme runs	Run records: one folder per design session in runs/ (start, record, list).

cheme

Plant design as code: validate, format, render and diff a plant repository.

cheme [OPTIONS] COMMAND [ARGS]...

name	kind	type	required	default	help
--version	option	flag			Print the cheme-as-code and pyDEXPI versions and the interpreter, and exit.

Subcommands: init, validate, fmt, render, diff, show, check, status, impact, maturity, issues, cause-effect, serve, export-static, git-setup, schema, migrate, copy, fragment, run, regen, move, rename, set, sheet, process, layout, assess, datasheets, skills, chemistry, control, safety, experiments, data, runs

cheme init

Create a plant repository from the template (folders, schemas, skills, CI).

cheme init [OPTIONS] [directory]

name	kind	type	required	default	help
directory	argument	path		.	Where to create the repository.
--name	option	str		``	Plant name for plant.yaml.
--force / --no-force	option	flag			Overwrite existing files.

cheme validate

Validate every P&ID and list, and the lists against the P&IDs.

A sheet declared generated in plant.yaml is rebuilt from its generator in a scratch file and compared (cheme.generated-sheet-drift); --skip-generators leaves that out.

cheme validate [OPTIONS] [paths]...

name	kind	type	required	default	help
paths	argument	path			Only these files.
--json	option	flag			Machine-readable report.
--fail-on	option	error | warning | info		error	Lowest level that fails: error, warning or info.
--quiet / -q	option	flag			Print only failures.
--skip-generators	option	flag			Do not run generators to check generated sheets for drift.
--ignore	option	str			Leave out these rules (ids or globs); also for --json and --fail-on.
--only	option	str			Report only these rules (ids or globs).

cheme fmt

Write P&IDs in canonical form: readable ids on every object, DEXPI 2.0.1, stable order.

cheme fmt [OPTIONS] [paths]...

name	kind	type	required	default	help
paths	argument	path			P&ID files (default: all).
--check / --no-check	option	flag			Only report files that would change.

cheme render

Render P&IDs to SVG or PNG: build/<sheet>(@<rev>)(_crop-X0-Y0-X1-Y1)(_inline).svg|png.

cheme render [OPTIONS] [paths]...

name	kind	type	required	default	help
paths	argument	path			P&ID files (default: all).
--png	option	flag			Write PNG (needs the png extra).
--scale	option	float range			PNG pixels per sheet millimetre (default 4; 8 with --tiles).
--crop	option	str			Only this sheet region: X0,Y0,X1,Y1 in sheet mm, origin top left, Y down.
--rev	option	str			Render the sheets as of this git revision.
-o	option	path			Output file (one input).
--tiles	option	str			Overlapping tiles covering the sheet's whole viewBox (which can include content left of or above the frame), C columns by R rows (e.g. 4x3), or auto: build/<sheet>_tile-<row letter><column>.svg|png and build/<sheet>_tiles.json.
--findings	option	flag			Mark the drawing-geometry findings on it.
--emphasise-inline	option	flag			Draw valves, inline fittings and actuators 1.5 times larger with bolder lines, for reading a full-sheet render (a review aid; the XML is unchanged): build/<sheet>..._inline.svg|png.

cheme diff

Compare P&IDs between two git revisions; write an HTML report with changes highlighted.

--json or --list print the result instead; the HTML report is then written only with --html. --list groups the changes by the plant item they belong to; --flat lists every object. A sheet that does not load or draw is named with the reason; the others are still compared, and the exit code is 1.

cheme diff [OPTIONS] [old_rev] [new_rev]

name	kind	type	required	default	help
old_rev	argument	str		HEAD	Old revision, e.g. main.
new_rev	argument	str			New revision (default: the working tree).
--path	option	path			Only these sheets.
--html	option	path			HTML report (default: build/diff.html).
--json	option	flag			Print every sheet's changes as JSON.
--list	option	flag			Print one line per changed plant item.
--flat	option	flag			With --list: one line per changed object (0.2).
--verbose / -v	option	flag			With --list: each change under its item.

cheme show

Where an object is: its block, owner, drawing, references and what references it.

cheme show [OPTIONS] {ID_OR_TAG}

name	kind	type	required	default	help
ID_OR_TAG	argument	str	yes		An XML id or a tag.
--sheet	option	path			Only this sheet (default: every sheet).
--layout	option	flag			Measured plot-plan distances to the nearest items, buildings, roads, racks and the site boundary.

cheme check

Run the engineering-step gates in checks/checks.yaml (--write signs off a stage).

cheme check [OPTIONS]

name	kind	type	required	default	help
--stage	option	str			Gates of every stage up to this one (plant.yaml order).
--only-stage	option	flag			With --stage: only that stage's gates.
--write / --no-write	option	flag			Sign off the stage: record it in checks/status.json (refused unless the tree is clean and every gate passes).
--review	option	path			With --write: the review record the sign-off rests on.
--by	option	str			With --write: who signs off (default: git user.name).

cheme status

Where the repository stands: sheets, findings by sheet, gates, sign-offs, review findings.

cheme status [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			Machine-readable summary.
--rev	option	str			The repository as committed at this revision (a temporary worktree).
--gates / --no-gates	option	flag		True	Evaluate the gates (they may run commands).

cheme impact

Artifacts whose sources changed since they were reviewed or generated: review needed.

Review records whose scope or basis changed since, generated sheets whose process YAML changed, stale calculated values. Read-only, info only, exit 0: nothing is regenerated.

cheme impact [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			Machine-readable findings.

cheme maturity

Which calculation maturity level the evidence supports, requirement by requirement.

Levels: sourced process description, balanced conceptual PFD, converged predictive simulation, preliminary/basic engineering package. The design maturity a process.yaml claims, artifact completeness and review status are reported beside them and never raise a level; no command sets the review status (docs/concepts/glossary.md).

cheme maturity [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			The machine-readable report.
--write / --no-write	option	flag			Also write the report to checks/maturity.json.
--require	option	str			Exit 1 unless this calculation maturity level is reached: sourced, balanced, predictive, engineering.

cheme issues

The plant's issues (plant.yaml issues:) as engineering items: actions, holds, technical queries and design changes, by the gate they must close before. Exits 2 when they cannot be read (no GitHub token or network).

cheme issues [OPTIONS]

name	kind	type	required	default	help
--all	option	flag			Closed issues too.
--json	option	flag			As JSON.

cheme cause-effect

Write the cause-and-effect matrix of control/cause-effect.yaml (HTML and CSV).

cheme cause-effect [OPTIONS]

name	kind	type	required	default	help
-o	option	path			HTML file (default: build/cause-effect.html).

cheme serve

Serve the repository read-only as a JSON API (and the viewer, with --ui).

In a folder of plants (process-library's root, or a folder of plant repositories) it serves each plant under /p/<id>/ and lists them at /api/plants (ADR 0034).

cheme serve [OPTIONS]

name	kind	type	required	default	help
--repo	option	path			Plant repository (default: the one around the cwd), or a folder of plants (a process library): then every plant is served under /p/<id>/.
--library	option	flag			Serve every plant below the folder (at most 3 levels down), also when it lies inside a plant. Without it, a folder with no plant.yaml above it but plants below it is served as a library.
--host	option	str		127.0.0.1	Address to bind. A non-local address (e.g. 0.0.0.0) serves the repository to the network; requests must still name a local host or an --allow-host name.
--port	option	int		8765	Port to listen on.
--ui	option	path			Built viewer to serve at / (a dist directory).
--allow-host	option	str			Also accept requests for this host name or address in the Host header (repeatable), e.g. the machine's LAN address with --host 0.0.0.0. Only 127.0.0.1, localhost, ::1 and a non-wildcard --host are accepted otherwise (against DNS rebinding); no wildcard.
--allow-remote-checks / --no-allow-remote-checks	option	flag			Let the viewer run the checks (POST /api/checks/run, which runs the command checks of checks/checks.yaml) although --host is not a local address.

cheme export-static

Write the viewer's API as static files: a read-only snapshot the viewer opens without a server (manifest.json + data/).

cheme export-static [OPTIONS] {out}

name	kind	type	required	default	help
out	argument	path	yes		Output directory: new, empty, or a previous snapshot (replaced). Not the plant repository, a directory containing it, or one inside it other than below build/ (e.g. build/snapshot).
--repo	option	path			Plant repository (default: the one around the cwd).
--rev	option	str			Export this git revision (default: the working tree).
--max-file-mb	option	float range		2.0	Leave out files larger than this (MB).
--exclude	option	str			Leave out paths matching this glob (repeatable; ** spans directories, e.g. reviews/** or PROGRESS.md).

cheme git-setup

Show P&IDs as review text in git diffs, and check them before each commit.

cheme git-setup [OPTIONS]

name	kind	type	required	default	help
--hook / --no-hook	option	flag		True	Install the pre-commit hook.

cheme schema

Print the JSON Schema of a YAML file kind, or list the kinds.

cheme schema [OPTIONS] [kind]

name	kind	type	required	default	help
kind	argument	str			plant, equipment, lines, site_plan, ... (none: list the kinds)

cheme migrate

Bring the YAML inputs to the schema versions this cheme-as-code reads: a file without schema: gets schema: cheme/<kind>/v1 as its first key, its comments kept, and a v1 LCI dataset in lca/datasets/ (schema: cheme/lca/v1) gets schema: cheme/lci_dataset/v2. Exits 1 when --check finds a file to change or a file cannot be migrated.

cheme migrate [OPTIONS]

name	kind	type	required	default	help
--check	option	flag			Only list the files that would change (default).
--write	option	flag			Rewrite the files.

cheme copy

Duplicate an object with its parts and drawing, renaming every id (e.g. a spare pump).

cheme copy [OPTIONS] {ID}

name	kind	type	required	default	help
ID	argument	str	yes		The object to copy (id or tag).
--prefix	option	str	yes		Id of the copy; every id inside it is derived from it.
--tag	option	str			Tag of the copy (TagName, ...).
--offset	option	str		0,0	Move the drawing by DX,DY (sheet mm, Y down).
--sheet	option	path			The sheet (default: the one holding ID).
--into	option	str			Put the copy into this object.
--print	option	flag			Print the copied XML; change no file.
--rename	option	str			An inner tag of the copy (data, labels, ids).
--renumber	option	str			A number in the copy's inner tags, labels, ids.
--set	option	str			On every copied object having PROP.
--sense-at	option	str			Instrument: sensing location (and its node).
--connect	option	str			Connect a copied pipe's open end.

cheme fragment

Print a ready-to-paste XML fragment (conceptual part and drawing part).

cheme fragment [OPTIONS] {kind} [name]

name	kind	type	required	default	help
kind	argument	str	yes		node, pipe, valve, tee, segment, label, equipment, nozzle, instrument, signal, offpage, line, note or shape.
name	argument	str			shape: the catalogue name, e.g. COLUMN_SHAPE.
--id	option	str			Id of the new object; its parts' ids derive from it.
--dn	option	str			Nominal diameter, e.g. 80.
--standard	option	str		Din2448	DN standard: Din2448, Iso6708 or none.
--from	option	str			pipe/segment: where it starts (ITEM:NODE); signal: the source instrument.
--to	option	str			pipe/segment: where it ends (ITEM:NODE); signal: the operated valve(s) or instrument(s) it goes to, comma-separated for a trip.
--points	option	str			pipe/signal/instrument (measuring line): the bends between the two node positions, x,y;x,y (default: routed).
--class	option	str			valve: GlobeValve (default), BallValve, GateValve, Silencer, ...; equipment: the DEXPI class, e.g. CentrifugalCompressor.
--at	option	str			Position X,Y (sheet mm, Y down).
--rotation	option	float		0.0	valve/tee: flow direction, degrees (0: left to right); equipment/offpage: symbol rotation, degrees.
--branch-dn	option	str			tee: DN of the branch node.
--branch	option	str			tee: side the branch leaves (absolute): up, down, left or right (default: left of the flow, up for --rotation 0).
--for	option	str			label: the object it shows.
--attribute	option	str			label: the attribute shown, e.g. TagName.
--tag	option	str			valve: PipingComponentNumber; tee: its tag; equipment: K-1101; instrument: PIC-4101; nozzle: SubTagName.
--number	option	str			segment: SegmentNumber (S2); line: LineNumber; signal: ActuatingSystemNumber.
--like	option	str			segment: copy fluid, class and DN data from this segment (--sheet).
--sheet	option	path			Take shape ids, node positions and values from it.
--shape	option	str			equipment: symbol name (default: by --class).
--scale	option	str			equipment: symbol scale S or SX,SY; valve: S.
--nozzle	option	str			equipment: a nozzle, offset from the symbol centre; no DN: an instrument connection. The stub points away from the nearer side of the symbol, or DIR (left, right, up, down; N1:-10,40::left without DN). Repeat it.
--marker	option	str			equipment: a marker symbol inside it (COLUMN_PACKED_BED_SHAPE, COLUMN_TRAY_SHAPE), dy from the symbol centre, ScaleY SY (ScaleX: the equipment's), m: mirrored. Repeat it.
--label-at	option	str			equipment: tag label position X,Y.
--on	option	str			nozzle: the equipment's id.
--sense	option	str			instrument: what it measures (nozzle, segment, inline item), and where the measuring line starts.
--field / --central	option	flag			instrument: field or central bubble.
--fail-action	option	str			signal to a valve: FC, FO or FL.
--in	option	flag			offpage: flow comes in from another sheet.
--out	option	flag			offpage: flow goes out to another sheet.
--note	option	str			offpage: text beside the arrow.
--note-at	option	str			offpage: position of --note.
--text	option	str			note: the text.
--fluid	option	str			line: FluidCode.

cheme run

Run a Python script in cheme-as-code's interpreter, where pyDEXPI and chemeascode import.

Usage: cheme run [--fmt SHEET] SCRIPT [ARGS...]. Options go before SCRIPT: everything after it goes to the script (sys.argv[1:]); its exit code is cheme-as-code's. The working directory is the repository root, so relative paths mean the same wherever you start it. Use this instead of python3 or uv run python: neither has pyDEXPI in a plant repository.

cheme run [OPTIONS] {script}

name	kind	type	required	default	help
script	argument	path	yes		The Python file to run, e.g. scripts/draft_P-02.py.
--fmt	option	path			After a successful run, run cheme fmt on this sheet (repeatable).

cheme regen

Rebuild generated sheets: frame -> cheme run SCRIPT FILE -> fmt.

A sheet is generated when plant.yaml gives it generator: {script, frame}; it is then not edited by hand. regen refuses a sheet that differs from what its generator (as committed) makes, naming the items edited by hand, unless --force. --detach removes the generator: the XML becomes the source and the script provenance only.

cheme regen [OPTIONS] [sheets]...

name	kind	type	required	default	help
sheets	argument	str			Sheet ids or files (default: every sheet with a generator).
--force / --no-force	option	flag			Overwrite a sheet edited by hand since its last generation.
--detach	option	flag			Make the sheets hand-maintained: remove their generator.
--list	option	flag			Print which sheets are generated; change nothing.

cheme move

Move an object's drawing: symbol, node positions, labels; attached lines follow.

cheme move [OPTIONS] {ID}

name	kind	type	required	default	help
ID	argument	str	yes		The object to move (id or tag).
--offset	option	str	yes		DX,DY in sheet mm (Y down), e.g. -8,12.
--sheet	option	path			The sheet (default: the one holding ID).
--print	option	flag			Dry run: print the change as a diff; write nothing.
--no-fmt	option	flag			Do not run cheme fmt after writing.

cheme rename

Rename an id and every reference to it, or a tag and the labels that show it.

cheme rename [OPTIONS] {old} {new}

name	kind	type	required	default	help
old	argument	str	yes		The id (or a tag) to rename.
new	argument	str	yes		The new id (or tag).
--children / --no-children	option	flag			Also rename the ids derived from it: OLD_* -> NEW_*.
--ids / --no-ids	option	flag			Renaming a tag: also rename the ids that spell it (HV4146 -> HV4147).
--tag	option	flag			NEW is the new tag of the object OLD names (by id): retags one of two objects sharing a tag (an instrument's sensor and actuator follow).
--sheet	option	path			The sheet (default: the one holding ID).
--print	option	flag			Dry run: print the change as a diff; write nothing.
--no-fmt	option	flag			Do not run cheme fmt after writing.

cheme set

Set data values of an object (typed as DEXPI defines them) and the labels showing them.

cheme set [OPTIONS] {ID} {PROPERTY=VALUE...}

name	kind	type	required	default	help
ID	argument	str	yes		The object (id or tag).
PROPERTY=VALUE...	argument	str	yes		DEXPI attribute names and values (FluidCode=Nitrogen, NominalCapacityVolume='26 MetreCubed', Stroke.Width=0.5); PROPERTY= removes it.
--sheet	option	path			The sheet (default: the one holding ID).
--print	option	flag			Dry run: print the change as a diff; write nothing.
--no-fmt	option	flag			Do not run cheme fmt after writing.

cheme sheet

Start a P&ID sheet with a real title block, or change its title block.

cheme sheet [OPTIONS] COMMAND [ARGS]...

Subcommands: new, meta

cheme sheet new

Write a new sheet from the template, its title block filled from plant.yaml.

Sets the Model name and uri, the diagram Name and the MetaData (drawing and sheet number, title, project, client, site, revision, dates, status "for review"; approver, approval date and checker stay empty) with their title-block texts, and adds the sheet to plant.yaml (or marks a planned one drafted).

cheme sheet new [OPTIONS] {ID}

name	kind	type	required	default	help
ID	argument	str	yes		Sheet id, e.g. P-05.
--title	option	str		``	Sheet title (DrawingName); default: its plant.yaml title.
--file	option	str			Path from the repository root (default pid/<ID>.xml).
--revision	option	str			RevisionNumber (default: plant.yaml project.revision).
--force / --no-force	option	flag			Overwrite an existing file.

cheme sheet meta

Show or set title-block values: the MetaData and the title-block texts together.

cheme sheet meta [OPTIONS] {SHEET} [KEY=VALUE...]

name	kind	type	required	default	help
SHEET	argument	str	yes		Sheet id or file.
KEY=VALUE...	argument	str			MetaData property (RevisionNumber) or short name (Revision, Status, Title, Approver, Checker, Client, Site, Project, ...), or ModelName, ModelUri, DiagramName. An empty value empties it. None: print the title block.

cheme process

Process design: BFD and PFD generated from process/materials.yaml and process/flowsheet.yaml.

cheme process [OPTIONS] COMMAND [ARGS]...

Subcommands: build, streams, reactors, crystallizers, reactor-uncertainty, reactor-hazards, reactor-choice, adsorption, cell, duty, stack, energy, pinch, furnace, props, propagate, simulate, phase, extract, methods, electrolyte, assay, layout

cheme process build

Generate the BFD and PFD sheets (DEXPI 2.0.1 ProcessModels with a drawing).

Refuses while the flowsheet has errors, and over a sheet edited by hand unless --force.

cheme process build [OPTIONS] [sheets]...

name	kind	type	required	default	help
sheets	argument	str			Sheet ids or files (default: every bfd/pfd sheet).
--force / --no-force	option	flag			Overwrite a sheet edited by hand since it was generated.

cheme process streams

The stream table (value, unit, provenance per property) and the block balances.

cheme process streams [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: the default case).
--md	option	flag			Markdown table.
--csv	option	flag			CSV.
--json	option	flag			Streams, balances and provenance as JSON.

cheme process reactors

Each reactor: model, extents, heats of reaction, ln K and ln(Q/K) at the outlet, and the duty H(out) - H(in) on the elements basis (ideal gas).

cheme process reactors [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: the default case).
--json	option	flag			As JSON.

cheme process crystallizers

Each crystallizer and filter or centrifuge (crystallizer:, solid_separation:): its basis (solubility-equilibrium, imposed-yield or solid-split), crystals, recovery, the outlets as dry solid and liquid, the balances and the duty. Exit 1 on an error finding.

cheme process crystallizers [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: the default case).
--json	option	flag			As JSON.

cheme process reactor-uncertainty

The spread of NODE's conversion, outlet flows and temperatures from the uncertainty of its reaction data, by Monte Carlo through its own reactor model. Only what the model reads is sampled (rate A and Ea; K with reverse: equilibrium; heat_of_reaction for a liquid adiabatic tank or batch; conversion for a conversion reactor; for a selectivity: step, the key's conversion and the selectivities from selectivity.uncertainty, evaluated as propagate does); other uncertainties are listed as not used, with the reason. Writes nothing. Exit 2 when it cannot be sampled or nothing it uses is uncertain.

cheme process reactor-uncertainty [OPTIONS] {node}

name	kind	type	required	default	help
node	argument	str	yes		The reactor block or step.
--samples	option	int range		200	Monte Carlo draws.
--seed	option	int		1	Seed of the generator.
--case	option	str			Case id (default: the default case).
--json	option	flag			As JSON.

cheme process reactor-hazards

Thermal screening of each reactor: adiabatic temperature rise, MTSR after a cooling failure, MTT, TD24, Stoessel's criticality class and, for cooled beds, the hot spot's sensitivity to the coolant temperature (domains/safety/reaction_hazards.py). Screening values only. Exit 1 on a warning (class 3 to 5, an MTSR above MTT or TD24, a sensitive bed).

cheme process reactor-hazards [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: the default case).
--json	option	flag			As JSON.

cheme process reactor-choice

Which ideal reactor suits NODE's reactions: the yield and selectivity of PRODUCT against the key's conversion for a plug-flow reactor, one stirred tank and --tanks tanks in series (isothermal, at the inlet's temperature and pressure), and the plug flow's instantaneous selectivity. --svg writes the plot. Exit 2 when it cannot be compared.

cheme process reactor-choice [OPTIONS] {node}

name	kind	type	required	default	help
node	argument	str	yes		The reactor block or step.
--product	option	str	yes		The desired product.
--key	option	str			The converted component (default: first fed).
--tanks	option	int range		3	Tanks in series to compare.
--case	option	str			Case id (default: the default case).
--svg	option	path			Write the plot to this file.
--region-svg	option	path			Also compute the candidate attainable region and write its plot here.
--json	option	flag			As JSON.
--csv	option	flag			The points as CSV.

cheme process adsorption

Each adsorber of process/adsorption.yaml: core's equilibrium TSA cycle (breakthrough, recoveries, regeneration heat and purge, adsorbent replacement) from sourced isotherms, or an external performance model's results checked against its operating range; and whether the split process/simulation.yaml gives the simulator follows from it. An isotherm outside its stated range and a calculated PSA are refused. Writes nothing. Exit 1 on a warning or error, 2 for an unknown case.

cheme process adsorption [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: every case).
--json	option	flag			As JSON.

cheme process cell

The Faraday balance of each electrochemical cell of process/cells.yaml: extents, products per outlet, charge and element balances, electric power, thermoneutral voltage, heat duty (Q = P - extent dH, electricity counted once) and kWh per kg of product, from the specified current, current efficiency and voltage. Refuses a cell without them. Writes nothing. Exit 1 on an error finding (a refused or inconsistent cell), 2 without process/cells.yaml.

cheme process cell [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			As JSON.

cheme process duty

The dynamic duties of process/duties.yaml: each cell's current or power schedule against the anolyte reservoir (voltage, efficiencies and WDE from the measured series or the specified cell, never extrapolated), the pH, inventory, product, charge, energy, dosing and bleed over time, the time to the stated pH threshold and the time outside the measured range. A duty a stack of process/cells.yaml names (equipment.duty) runs at the arrangement that stack is sized at, as in cheme process stack (cheme.duty-arrangement). --compare ID,ID --co KG compares duties at equal delivered product (the duties are named there, not as IDS; --plot then draws the compared transients). Writes nothing but the plots. Exit 1 on an error finding, 2 without process/duties.yaml, for an unknown duty id or for IDS given with --compare.

cheme process duty [OPTIONS] [ids]...

name	kind	type	required	default	help
ids	argument	str			Duty ids (default: every duty).
--json	option	flag			As JSON.
--plot	option	path			Write the time plots and operating maps (SVG) into this directory.
--compare	option	str			ID,ID: compare these duties at equal delivered product (--co).
--co	option	float			With --compare: the delivered product, kg.
--tolerance	option	float		0.001	With --compare: relative tolerance on the delivered product.

cheme process stack

The stacks of the stacks: section of process/cells.yaml sized for their product duty at the cell's operating point (never outside its measured or declared range): active area, cells, cells per stack from the voltage limit, stacks, stack voltage and current, DC power, output and its margin, kWh per kg, the stack envelope; and the skid equipment inputs from the duty named, run at the sized arrangement over its horizon: reservoir and tank dimensions (a dosing or bleed tank over its refill or emptying interval), the recirculation and dosing pumps' flows, and with the duty's reservoir.size: to_horizon the reservoir volume that reaches its horizon. Without --write it writes nothing and reports what differs from the equipment list, the datasheets and process/duties.yaml (cheme.stack-outdated); --write writes them (a reservoir volume first, then the tanks from the duty at it); with --write --json the report (as the files are then) carries written, each file written with its tags (or duty ids). Exit 1 on an error finding (process/cells.yaml not following its schema among them: its findings are printed), 2 without process/cells.yaml or stacks, or for an unknown id.

cheme process stack [OPTIONS] [ids]...

name	kind	type	required	default	help
ids	argument	str			Stack ids (default: every stack).
--json	option	flag			As JSON.
--write	option	flag			Write the calculated stack and tank dimensions into the equipment list, the pumps' inputs into their datasheets (never over a value written by hand) and a reservoir volume the duty asks for (reservoir.size: to_horizon) into process/duties.yaml.

cheme process energy

Energy balances of blocks and steps (stream enthalpies against energy streams), the energy audit of the recorded simulation run, and the boundary-completeness gaps (utilities, waste and emission terminals not declared). Exit 1 on a warning or error.

cheme process energy [OPTIONS]

name	kind	type	required	default	help
--case	option	str			Case id (default: the default case).
--json	option	flag			As JSON.

cheme process pinch

Whole-plant pinch targets of the case's recorded simulation run (process/simulation-run.json, from cheme process simulate --fill; the simulator is not run).

Prints the current utility demand, the theoretical minimum utility demand at dTmin and the theoretical energy-saving potential (their difference: not a saving), the hot-side, cold-side and shifted pinch (or a threshold problem), the utility-use pinch violations apart from cross-pinch transfer in existing exchangers, the thermal streams with class and quality, and the reconciliation against the run's duties. --write records the result in process/pinch-run.json. Exit 1 on an error (no current converged run, too few streams) or a reconciliation finding; 2 on invalid input; 3 when --write is refused (a reconciliation finding), nothing written.

cheme process pinch [OPTIONS]

name	kind	type	required	default	help
--dt-min	option	float	yes		The global minimum approach temperature dTmin, K (required).
--dt-min-source	option	str		ASSUMED	Where dTmin comes from (recorded; default: ASSUMED).
--case	option	str			Case id (default: the default case).
--include-constrained	option	flag			Include the constrained duties (reactors) in the targets, as a sensitivity.
--json	option	flag			As JSON.
--md	option	flag			As Markdown.
--write	option	flag			Record the targets in process/pinch-run.json.

cheme process furnace

Energy basis of the solid-reaction furnaces of process/furnaces.yaml (lime calcination): element balance, product and off-gas composition, reaction enthalpy and the furnace energy closure from condensed-phase NASA data (extra cantera), or a labelled specific-energy fallback. Exit 1 on a warning or error.

cheme process furnace [OPTIONS] [furnace_id]

name	kind	type	required	default	help
furnace_id	argument	str			Furnace id (default: every furnace).
--json	option	flag			As JSON.

cheme process props

Stream properties (phase, vapour fraction, molar mass, density, enthalpy) through thermo (extra props), with the process's property package, inside its validated range only.

--fill never overwrites given or assumed values; a state outside the range is refused and not written, and values calculated earlier stay and are reported stale. --fill exits 3 when it refused a stream (out of range, unsupported, failed or a contradicted phase) and wrote the rest. --check exits 1 on a warning (stale, differing or out-of-range values). --coverage names the package (materials.yaml property_method.package, or the one covering the most streams) and per stream whether it covers its components and state or why not, nearest other package included; it needs no thermo and exits 0 unless the recorded package is unknown. --json prints the run as an adapter result (cheme/adapter-result/v1) per case, same exit codes (with --coverage: the coverage as JSON).

cheme process props [OPTIONS]

name	kind	type	required	default	help
--fill	option	flag			Write calculated values and the method.
--check	option	flag			Recalculate and report; write nothing.
--coverage	option	flag			List the property package and, per stream, whether it covers it or why not; write nothing.
--case	option	str			Case id (default: the default case).
--json	option	flag			The run as an adapter result (JSON).

cheme process propagate

Derive missing compositions and flows through steps their streams fully determine: pass-through steps (compressors, coolers, heaters, pumps, filters), mixers, splitters (split: fractions per outlet port, else the mass balance remainder).

Separators and reactors are only checked, never solved. Derived values are calculated with an input fingerprint; given and assumed values are never overwritten. --check exits 1 on a warning (stale propagated values). --json prints the run as an adapter result (cheme/adapter-result/v1) per case, same exit codes.

cheme process propagate [OPTIONS]

name	kind	type	required	default	help
--fill	option	flag			Write the derived compositions and flows.
--check	option	flag			Derive and report; write nothing.
--case	option	str			Case id (default: the default case).
--json	option	flag			The run as an adapter result (JSON).

cheme process simulate

Solve the flowsheet in DWSIM, or in IDAES where process/simulation.yaml says simulator: idaes (optional; CHEME_IDAES_PYTHON names a Python with idaes-pse): feeds from process/flowsheet.yaml, simulator, property package and step specifications from process/simulation.yaml.

--fill writes temperature, pressure, flow and composition of every simulated stream that lacks them (and the duty of energy streams) as calculated with a fingerprint; given and assumed values are never overwritten, a deviation from them is a warning. --fill also records the run (convergence, unit results, stream values, input fingerprints) in process/simulation-run.json; --check reports a record made from other inputs. Exits 1 on an error (unmapped step, missing spec or feed, no simulator) and, with --check, on a warning (stale or deviating values, a stale record); 3 when the simulation did not converge (no stream value written, the run recorded).

cheme process simulate [OPTIONS]

name	kind	type	required	default	help
--fill	option	flag			Write the simulated values streams lack.
--check	option	flag			Simulate and report; write nothing.
--case	option	str			Case id (default: the default case).
--save	option	path			Also save the DWSIM flowsheet here (.dwxmz; DWSIM only).

cheme process phase

Binary interaction parameter coverage and, with --temperature and --pressure, a flash (phase identification) through thermo (extra props). No solids, electrolytes or pseudocomponents; no application range claimed.

Each pair is sourced (with its source), missing (the calculation uses zero) or accepted_zero (--accept-zero). Exits 2 on invalid input, 1 with --check when a pair is missing.

cheme process phase [OPTIONS] {components}...

name	kind	type	required	default	help
components	argument	str	yes		Component names or CAS numbers.
--model	option	str			Equation of state: Peng-Robinson, Redlich-Kwong-Soave, Redlich-Kwong.
--liquid	option	str			Activity-coefficient model: NRTL, Wilson, UNIFAC.
--vapor	option	str			With --liquid: an equation of state or ideal_gas.
--z	option	str			Overall mole fractions, comma-separated (mixtures).
--temperature	option	float			Flash temperature.
--temperature-unit	option	str		K	Unit of the temperature.
--pressure	option	float			Flash pressure (absolute).
--pressure-unit	option	str		bar(a)	Unit of the pressure.
--two-liquids	option	flag			Allow a second liquid (stated assumption).
--accept-zero	option	str			A pair A:B whose missing parameter is accepted as zero (repeatable).
--check	option	flag			Exit 1 when a binary parameter is missing.
--json	option	flag			The result as JSON.

cheme process extract

Liquid-liquid extraction from process/extraction.yaml: countercurrent equilibrium stages on NRTL (thermo, extra props) with a bundled or a sourced LLE system. Prints raffinate, extract, solute recovery, solvent loss, distribution and selectivity, closure, and whether the result is inside a benchmarked range (qualified). Solvent regeneration is a separate unit. Writes nothing. Exit 1 when a unit is refused (e.g. no two-phase region), fails or does not converge, 2 without the file or for an unknown unit.

cheme process extract [OPTIONS] [unit]

name	kind	type	required	default	help
unit	argument	str			One unit of process/extraction.yaml (default: all).
--systems	option	flag			List the bundled LLE systems and exit.
--json	option	flag			{lines, results: [adapter result per unit]}.

cheme process methods

The supported-method contract (methods.yml): per thermodynamic model the selection tree can recommend, whether cheme process phase and cheme process simulate run it, else its explicit substitute; and the coverage decisions for electrolytes, solids and pseudocomponents.

cheme process methods [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			The table as JSON.

cheme process electrolyte

Mean ionic activity coefficient, osmotic coefficient, water activity and vapour pressure, and relative enthalpy of a qualified aqueous electrolyte system (electrolytes.yml; today NaCl-H2O) through Pitzer's equations in Cantera (extras cantera and props).

Inside the claimed range only: anything else (another salt or ion, a charge imbalance, a state outside the range, a boiling solution) is refused with exit code 3. Exit 2 on invalid input or a missing extra.

cheme process electrolyte [OPTIONS]

name	kind	type	required	default	help
--temperature	option	float	yes		Temperature.
--pressure	option	float	yes		Pressure (absolute).
--molality	option	float			Molality of the salt, mol per kg of water.
--mass-fraction	option	float			Mass fraction of the salt in the solution.
--species	option	str			Solute species molalities, e.g. 'Na+=1,Cl-=1'.
--system	option	str		NaCl-H2O	The electrolyte system.
--temperature-unit	option	str		K	Unit of the temperature.
--pressure-unit	option	str		bar(a)	Unit of the pressure.
--json	option	flag			The result as JSON.

cheme process assay

Characterise the petroleum assays of process/assays.yaml into pseudocomponents (Riazi-Daubert, Lee-Kesler, Lastovka-Shaw; D86 converted to TBP by API 3A1.1) and print them with their identities and the characterisation fingerprint. An insufficient assay or one outside the correlations' range is refused (no pseudocomponents).

--pressure adds the bubble and dew temperatures, --temperature with --pressure a Peng-Robinson flash (kij = 0), each with why it lies outside the benchmarked claim. --fill records the characterisations in process/assay-characterization.json; --check reports a record made from other inputs or another method version. Exits 1 on an error (with --check also on a warning), 2 on a wrong call.

cheme process assay [OPTIONS] [ids]...

name	kind	type	required	default	help
ids	argument	str			Assay ids (default: all).
--fill	option	flag			Record the characterisations.
--check	option	flag			Report a missing or stale record; write nothing.
--prune	option	flag			With --fill: drop entries of removed assays.
--temperature	option	float			Flash temperature, K (with --pressure).
--pressure	option	float			Flash pressure, kPa(a): bubble and dew points, or a flash at it.
--json	option	flag			The result as JSON.

cheme process layout

Print the generated layout of a view; --pin writes it into flowsheet.yaml layout.<view> so later builds keep those positions.

cheme process layout [OPTIONS] [ids]...

name	kind	type	required	default	help
ids	argument	str			Object ids to pin (default: every node).
--view	option	str		bfd	bfd or pfd.
--pin	option	flag			Write the positions as overrides into the flowsheet.

cheme layout

3D layout: the plant model generated from layout/*.yaml, the lists and the P&IDs.

cheme layout [OPTIONS] COMMAND [ARGS]...

Subcommands: build, route, pin, render, report, cables, lifting

cheme layout build

Derive layout/plant-model.yaml: equipment parts from type and dimensions, nozzles from the P&IDs (point, direction, DN, lines), maintenance envelopes, steel structures, platforms, zones, the site geometry and a pipe route for every P&ID segment.

Prints the build's findings and the model rules (clashes, headroom, access, rack fill). Refuses while a layout file is invalid, and over a model edited by hand unless --force. A rebuild from the same inputs is byte-identical.

cheme layout build [OPTIONS]

name	kind	type	required	default	help
--check	option	flag			Write nothing; exit 1 when the model is not up to date.
--force / --no-force	option	flag			Overwrite a model edited by hand since it was built.

cheme layout route

Rebuild the model, routing only the given lines (the other pipes are kept from the committed model); without --line, the same as layout build.

cheme layout route [OPTIONS]

name	kind	type	required	default	help
--line	option	str			Route only this line (repeatable); keep the other routes.
--force / --no-force	option	flag			Overwrite a model edited by hand since it was built.

cheme layout pin

Copy a generated route (into piping/routing.yaml), a proposed platform (into the item's access: in layout/equipment-placement.yaml), a proposed stair or ladder (into its structure in layout/structures.yaml, or its rack in layout/site-plan.yaml) or the opening a wall penetration proposes (into its building's or enclosure's openings; a roof penetration is refused: route the line through a wall), proposed support steel (into pipe_supports) or a valve access platform (into structures, with the routes of its valves' pipes), so the next build keeps it.

cheme layout pin [OPTIONS] {target}

name	kind	type	required	default	help
target	argument	str	yes		A line number, pipe id, TAG/platform-N, a proposed stair or ladder (S/stair-to-L, S/ladder-to-L, RACK/ladder-start|end), a wall penetration (OWNER/LINE), proposed support steel (PIPE/SP-<arc>) or a valve access platform (VP-<valve tag>).

cheme layout render

Draw the committed model as build/layout-*.svg (or .png): the plot plan, a level plan or a section, with grid, tags, line numbers, north arrow, scale bar and title block.

cheme layout render [OPTIONS]

name	kind	type	required	default	help
--plan	option	flag			A plan (plot plan by default).
--level	option	str			Level plan of a structure or building level id.
--z	option	str			Level plan of the z range A,B (m).
--section	option	str			A section at x=VALUE or y=VALUE (m).
--depth	option	float		6.0	Section depth (m).
--crop	option	str			X0,Y0,X1,Y1: site m (a section: along, z).
--png	option	flag			PNG (extra png) instead of SVG.

cheme layout report

Routed and unrouted lines with the reasons, lengths per line, rack fill, the model rule findings, and the generated / pinned / declared counts of the committed model.

cheme layout report [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			The report as JSON.

cheme layout cables

The cable schedule of the committed model: every cable from, to, size, trays, length (Markdown by default).

cheme layout cables [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			The schedule as CSV.
--json	option	flag			The schedule as JSON.
--write	option	flag			Write electrical/cable-schedule.csv and .md.
--check	option	flag			Exit 1 when the written schedule is not current.

cheme layout lifting

The lifting study of the committed model: per heavy item its crane or pad, radius, load moment, capacity needed, lift and hook height, lane and laydown; checked against layout/crane-chart.yaml when there is one (Markdown by default).

cheme layout lifting [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			The study as CSV.
--json	option	flag			The study as JSON.
--write	option	flag			Write layout/lifting-study.csv and .md.
--check	option	flag			Exit 1 when the written study is not current.

cheme assess

Assessment: the material take-off of the 3D model, the estimates, the LCA and the TEA.

cheme assess [OPTIONS] COMMAND [ARGS]...

Subcommands: mto, estimate, equipment, lca, lca-scenarios, lca-cases, lci-estimate, tea, lca-search, lca-map

cheme assess mto

The material take-off of layout/plant-model.yaml (Markdown by default): pipe by DN, class and insulation; elbows, tees, reducers, flanges; valves; steel tonnage; grating, insulation and concrete.

cheme assess mto [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			CSV, one row per quantity.
--json	option	flag			Rows, totals, findings and provenance as JSON.
--write	option	flag			Write estimate/mto.csv and estimate/mto.md.
--check	option	flag			Exit 1 when estimate/mto.csv or .md is not current.

cheme assess estimate

The take-off priced with estimate/rates.yaml, then its factors (direct, indirect, contingency): Markdown by default. cheme holds no prices; every rate is the project's, with its status, source and year.

cheme assess estimate [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			CSV, one line per row and factor.
--json	option	flag			Lines, totals, findings and provenance as JSON.
--write	option	flag			Write estimate/estimate.csv and estimate/estimate.md.
--check	option	flag			Exit 1 when estimate/estimate.csv or .md is not current.

cheme assess equipment

The equipment estimate from published cost correlations (Turton et al., CAPCOST): the purchased and bare module cost of every listed item, escalated by the project's cost index in estimate/cost-basis.yaml (Markdown by default).

cheme assess equipment [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			CSV, one line per item.
--json	option	flag			Lines, totals, findings and provenance as JSON.
--write	option	flag			Write estimate/equipment-cost.csv and estimate/equipment-cost.md.
--check	option	flag			Exit 1 when estimate/equipment-cost.csv or .md is stale.

cheme assess lca

The LCA of the plant's operation per functional unit (Markdown by default): the flowsheet's feeds and products in the case of lca/lca.yaml, shared between the co-products, characterised with lca/factors.yaml (and the open profiles it includes) or a Brightway project. --inventory and --dataset alone need neither: the foreground is exported without a background database or credentials. Every exchange without a burden is a gap, never 0.

cheme assess lca [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			The impacts as CSV, one row per contributor.
--inventory	option	flag			The foreground inventory as CSV.
--dataset	option	flag			The unallocated LCI dataset (kind lci_dataset) as YAML.
--coverage	option	flag			How far the result can be read: the classification and its reasons, the gaps (exchanges without a burden), the sources, the limitations, the data quality rating (TfS DQR, PEF-style indicators, primary data share) and the water balance. Writes nothing.
--balance	option	flag			The mass, element (C, H, O, N, S, Cl, ...) and carbon-by-origin balance of the plant's LCI dataset and its energy closure: in, out and residual per quantity (kg per hour and %), the tolerance, and the exchanges excluded with their reasons (with --json: as JSON). Needs no factors or background. Writes nothing.
--json	option	flag			Inventory, impacts, findings and provenance as JSON.
--write	option	flag			Write lca/inventory.csv, results.csv, results.md, gaps.csv, the unallocated LCI dataset lca/dataset.yaml and the run manifest lca/manifest.json.
--check	option	flag			Exit 1 when one of the lca/ outputs is not current.
--pact	option	path			Write the product carbon footprint as a WBCSD PACT (Pathfinder) v2 ProductFootprint JSON to this file (needs carbon_accounting with pact: in lca/lca.yaml).
--strict	option	flag			Exit 1 when an exchange of the baseline or a scenario rests on an unpublished factor: an example factor, one ASSUMED, PROXY or ESTIMATE, or a gap (each named as the error cheme.lca-factor-status, with its factor status).
--case	option	str			With --dataset, --write or --check only: the unallocated LCI dataset of this flowsheet case instead of lca/lca.yaml's, its id, title and uuid naming the case; --write writes it to lca/datasets/<dataset id>-<case>.yaml, --check exits 1 when that file is stale. The case lca/lca.yaml studies is the same as no --case.

cheme assess lca-scenarios

One machine-readable record per LCA scenario (a summary by default): its stable id <dataset>/<case>/<scenario>, the design case, the unallocated foreground, each exchange's background in the scenario (geography, year, status, source; linked, proxy, missing), the characterised totals per functional unit with the allocation, the gaps, warnings, publication policy and input fingerprints. A regional scenario changes the background only; only a scenario naming its own flowsheet case changes the foreground.

cheme assess lca-scenarios [OPTIONS] [ids]...

name	kind	type	required	default	help
ids	argument	str			Scenario ids of lca/lca.yaml (baseline included); default: all.
--json	option	flag			The records as one JSON object, by path.
--write	option	flag			Write lca/scenarios/<id>.json, one per scenario.
--check	option	flag			Exit 1 when a scenario record is not current.

cheme assess lca-cases

Existing flowsheet cases side by side: each case's declared capacity and provenance, and each exchange per unit (kg, or kWh for an energy product) of the functional unit's product (unallocated). Only computed cases; nothing is scaled, and an exchange a case lacks is shown as missing, not 0.

cheme assess lca-cases [OPTIONS] [cases]...

name	kind	type	required	default	help
cases	argument	str			Flowsheet case ids; default: every case.
--json	option	flag			The comparison as JSON.

cheme assess lci-estimate

A Tier 0 (stoichiometric) or Tier 1 (sourced conceptual) LCI dataset for a process without a flowsheet (YAML by default): from the reactions of process/reactions.yaml, the formulas of process/materials.yaml and the estimate's yield and selectivity, with every default exchange marked assumed and its rule cited. Refuses a dataset whose element balance does not close.

cheme assess lci-estimate [OPTIONS] [spec]

name	kind	type	required	default	help
spec	argument	str		lca/estimate.yaml	The estimate (kind lci_estimate), from the repository root.
--write	option	flag			Write the dataset (default lca/datasets/<id>.yaml).
--check	option	flag			Exit 1 when the written dataset is not current.
--compare	option	flag			Compare the estimate with the repository's flowsheet (process/flowsheet.yaml), per kg of the product: the validation of a tier against a full design.
--case	option	str			The flowsheet case for --compare.
--json	option	flag			The dataset, its balance and the findings as JSON.

cheme assess tea

The techno-economic analysis of the plant (Markdown by default): the capital cost built up from the estimate, the operating cost and revenue of the flowsheet's feeds and products, the discounted cash flow (NPV, IRR, LCOP, payback), scenarios, what dominates the outcome and what would change the decision. cheme holds no prices; every one is the project's, in estimate/tea.yaml with its status, source and year.

cheme assess tea [OPTIONS]

name	kind	type	required	default	help
--csv	option	flag			The cash flow as CSV, one row per year.
--sensitivity	option	flag			Tornado and break-even values as CSV.
--json	option	flag			Premises, results, findings and provenance as JSON.
--write	option	flag			Write estimate/tea.md, tea-cashflow.csv, tea-sensitivity.csv and (with a Monte Carlo) tea-uncertainty.csv.
--check	option	flag			Exit 1 when one of the TEA outputs is not current.
--compare	option	str			Compare with the TEA at this revision (a temporary worktree): what the two do not share, then their results.

cheme assess lca-search

Find activities and flows in the Brightway project (extra lca), printed as the activity: or biosphere: mapping to paste into lca/lca.yaml.

cheme assess lca-search [OPTIONS] {text}

name	kind	type	required	default	help
text	argument	str	yes		Words of the activity's name or product.
--database	option	str			Only this database of the project.
--limit	option	int		20	At most this many.
--project	option	str			Brightway project; default brightway.project of lca.yaml.

cheme assess lca-map

The exchanges without a background (unmapped, or declared missing:) with candidate activities of the Brightway project (extra lca), numbered, with the query that found them. A dry run: it writes nothing unless --apply names the approved candidates, and then only their activity: mappings.

cheme assess lca-map [OPTIONS] [keys]...

name	kind	type	required	default	help
keys	argument	str			Only these exchanges (e.g. feed:POWER); default every open one.
--query	option	str			Search with this text instead of each exchange's name.
--database	option	str			Only this database of the project.
--limit	option	int range		5	At most this many per exchange.
--project	option	str			Brightway project; default brightway.project of lca.yaml.
--apply	option	str			KEY=N: write candidate N of KEY (as listed with the same --query, --database and --limit) as its activity: into lca/lca.yaml. Repeatable; needs --source.
--source	option	str			Why these mappings (their source: in lca/lca.yaml).

cheme datasheets

Equipment datasheets: drafts, calculated sizes and dimensions, outline drawings.

cheme datasheets [OPTIONS] COMMAND [ARGS]...

Subcommands: draft, fill, drawing

cheme datasheets draft

Write a draft equipment/datasheets/<TAG>.yaml for every listed item without one: type, service, design conditions and material from the equipment list, dimensions, nozzles from the P&IDs, inlet and outlet conditions from the flowsheet streams linked to it; every value with its source, unknowns TBD, draft: true. Existing datasheets are never touched.

cheme datasheets draft [OPTIONS]

name	kind	type	required	default	help
--tag	option	str			Only this item (repeatable).
--dry-run	option	flag			Print the drafts, write nothing.

cheme datasheets fill

Fill the TBD sizes of the draft datasheets (draft: true) with calculated values and their basis: exchanger duty, LMTD and area; pump head, NPSH available and power; compressor and expander stages and power; column shortcut stages; vessel volume and residence time. A value the datasheet gives is never overwritten; what cannot be calculated stays TBD with the reason. The main dimensions and the empty weight come from the equipment list and the layout model (--dimensions: only those, final datasheets included).

cheme datasheets fill [OPTIONS]

name	kind	type	required	default	help
--tag	option	str			Only this item (repeatable).
--param	option	str			NAME=VALUE: an assumption for every item without its own (repeatable), e.g. pump_efficiency=0.7; written ASSUMED. --list names them.
--case	option	str			Flowsheet case.
--dimensions	option	flag			Only the main dimensions and the empty weight the equipment list and the layout model state, into final datasheets too (never over a value written by hand).
--dry-run	option	flag			Print the filled drafts, write nothing.
--list	option	flag			List the parameters, the typical overall coefficients (u_service) and the pipe roughness rows, and exit.

cheme datasheets drawing

The outline drawing of each datasheet, equipment/datasheets/<TAG>.svg: an elevation and a nozzle orientation view with the stated main dimensions, the nozzle schedule, weights, materials and a title block, generated from the datasheet (never edited by hand). Without an option, lists each drawing's state (current, stale, missing).

cheme datasheets drawing [OPTIONS]

name	kind	type	required	default	help
--tag	option	str			Only this item (repeatable).
--write	option	flag			Write equipment/datasheets/<TAG>.svg where not current.
--check	option	flag			Exit 1 when a drawing is missing or not current.
--print	option	flag			Print the drawing (one --tag) and write nothing.

cheme skills

The bundled skills installed in .claude/skills/: status and safe updates.

cheme skills [OPTIONS] COMMAND [ARGS]...

Subcommands: status, update

cheme skills status

Each bundled skill file as installed: current, outdated (an earlier bundled version), modified (changed locally) or missing.

cheme skills status [OPTIONS] [names]...

name	kind	type	required	default	help
names	argument	str			Skills (default: all).
--json	option	flag			Machine-readable.

cheme skills update

Install the bundled version of missing and outdated skill files. A file changed locally is kept and reported (exit 1) unless --force.

cheme skills update [OPTIONS] [names]...

name	kind	type	required	default	help
names	argument	str			Skills (default: all).
--force / --no-force	option	flag			Also replace files changed locally (they are still reported).

cheme chemistry

Chemistry: component properties from databases and correlations.

cheme chemistry [OPTIONS] COMMAND [ARGS]...

Subcommands: props, thermo, kinetics, microkinetics, library, evidence

cheme chemistry props

Look up properties of one component; each line names the source of its value.

Without --source the web databases are asked (network). Exits 1 when a property was not found.

cheme chemistry props [OPTIONS] {component}

name	kind	type	required	default	help
component	argument	str	yes		Name or CAS number, e.g. ethanol, 64-17-5.
--property / -p	option	str	yes		Property to look up (repeat for several).
--temperature	option	float		298.15	Temperature.
--temperature-unit	option	str		K	Unit of --temperature.
--pressure	option	float		101325.0	Pressure.
--pressure-unit	option	str		Pa	Unit of --pressure.
--source	option	str			Ask only these sources (PubChem, NIST, OSHA, Thermo, GRAPPA).
--grappa-endpoint	option	str			Base URL of a running GRAPPA service (last resort).
--json	option	flag			Results as JSON.

cheme chemistry thermo

Ideal-gas formation data (dHf, S at 298.15 K, Cp/R polynomial) for process/materials.yaml.

Prints one ideal_gas: entry per component to paste into its components: line (extra props, offline). Exits 1 when a component is not known.

cheme chemistry thermo [OPTIONS] {components}...

name	kind	type	required	default	help
components	argument	str	yes		Names or CAS numbers.
--fit	option	str			LO,HI in K: fit the heat capacity to thermo over this range (beyond 1000 K).
--json	option	flag			Results as JSON.

cheme chemistry kinetics

Kinetics of an elementary gas-phase reaction from NIST (and RMG), reduced to one Arrhenius record, printed as a rate: entry for process/reactions.yaml (network).

The rate is a power law per volume on concentrations (kmol/m3), its A converted from SI moles. Exits 1 when no record is found.

cheme chemistry kinetics [OPTIONS]

name	kind	type	required	default	help
--reactant / -r	option	str	yes		ID=NAME or ID=SMILES, a reactant (repeat). A formula that is also a SMILES (CO, NO) is refused: give a name, [C-]#[O+], or smiles:CO.
--product / -p	option	str			ID=NAME or ID=SMILES, a product (repeat). A formula that is also a SMILES (CO, NO) is refused: give a name, [C-]#[O+], or smiles:CO.
--tmin	option	float			K; records must cover it (alone: the one temperature).
--tmax	option	float			K; records must cover it (alone: the one temperature).
--rmg-endpoint	option	str			Base URL of a running RMG service (else NIST only).
--json	option	flag			The record and rate as JSON.

cheme chemistry microkinetics

A reaction's elementary steps at steady state (mean field): coverages, TOF, degree of rate control of every step, apparent activation energy and reaction orders.

cheme chemistry microkinetics [OPTIONS] {reaction}

name	kind	type	required	default	help
reaction	argument	str	yes		A reaction id of process/reactions.yaml.
--temperature / -T	option	float	yes		K
--pressure / -p	option	str	yes		ID=BAR, a partial pressure (repeat).
--product	option	str			The component whose TOF is analysed (default: first product).
--cantera	option	flag			Check the steady state with Cantera (extra cantera).
--export	option	path			Write the mechanism as a Cantera YAML input to this file.
--json	option	flag			As JSON.

cheme chemistry library

The reaction library: industrial processes with their components (ideal-gas data, critical constants), reactions and typical catalysts. Without an id, list them; with one, show it; with --add, add it to process/materials.yaml, reactions.yaml and catalysts.yaml.

cheme chemistry library [OPTIONS] [entry]

name	kind	type	required	default	help
entry	argument	str			An entry id (none: list them).
--add	option	flag			Add the entry's components, reactions and catalysts.
--json	option	flag			As JSON.

cheme chemistry evidence

The route comparison of process/reaction-evidence.yaml: per route its sources and observations, the temperature, pressure, conversion, selectivity and yield reported (each kept apart), kinetics, conflicts, sources not read and gaps, then the model basis and the file's cheme.evidence-* findings. Exit 1 on an error finding.

cheme chemistry evidence [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			As JSON.

cheme control

Control: the loop register, operating sequences, narrative and control philosophy.

cheme control [OPTIONS] COMMAND [ARGS]...

Subcommands: report

cheme control report

Write the loop summary (control-loops.html, .csv) and the sequence report (control-sequences.html), with each file's review status and the control findings.

cheme control report [OPTIONS]

name	kind	type	required	default	help
-o	option	path			Folder to write to (default: build/).
--json	option	flag			Print the report as JSON.

cheme safety

Safety: HAZOP studies and their risk matrix.

cheme safety [OPTIONS] COMMAND [ARGS]...

Subcommands: hazop

cheme safety hazop

HAZOP studies in safety/hazop/<study-id>/study.yaml: start one, check it, look its bands up in the risk matrix, write its report, record its analysed inputs.

cheme safety hazop [OPTIONS] COMMAND [ARGS]...

Subcommands: init, check, rate, report, stamp, review

cheme safety hazop init

Start a draft study: safety/hazop/STUDY/study.yaml with the analysed inputs at their current revision and an empty worksheet. Refuses an existing study.

cheme safety hazop init [OPTIONS] {study}

name	kind	type	required	default	help
study	argument	str	yes		The study id: its directory under safety/hazop/.
--title	option	str		``	The study's title.
--input	option	str			A file or directory the study analyses (pid/P-01.xml, control/; repeat). A directory is recorded at the HEAD commit and must be committed.
--matrix	option	str			The risk matrix file (default safety/risk-matrix.yaml if it exists).

cheme safety hazop check

The findings of the studies (as in cheme validate) and the state of each: invalid, draft, assessed or reviewed. Read-only; exits 1 when a study has errors.

cheme safety hazop check [OPTIONS] [study]

name	kind	type	required	default	help
study	argument	str			The study id; none: every study.
--json	option	flag			As JSON.

cheme safety hazop rate

Look up the band of each rating (category, severity, likelihood) in the study's risk matrix. Without --write nothing is written and the command exits 1 when a band is missing, stale or edited, or a rating cannot be looked up. --write records the bands (with the matrix and input fingerprint) and changes nothing else in the file.

cheme safety hazop rate [OPTIONS] {study}

name	kind	type	required	default	help
study	argument	str	yes		The study id: its directory under safety/hazop/.
--target	option	str			Only this node, deviation or scenario id (default: the whole study).
--write	option	flag			Record the bands (default: only check them).
--json	option	flag			As JSON.

cheme safety hazop report

Write the study's report to build/hazop/STUDY.md and .html: state, analysed inputs, matrix, unresolved items, the worksheet with every rating and band, safeguards, actions, review status.

cheme safety hazop report [OPTIONS] {study}

name	kind	type	required	default	help
study	argument	str	yes		The study id: its directory under safety/hazop/.
-o	option	path			Directory (default build/; files in hazop/).

cheme safety hazop stamp

Show the analysed inputs at their recorded and current revisions; --write records the current ones (sha256 of each file). Only once the team has reassessed what changed.

cheme safety hazop stamp [OPTIONS] {study}

name	kind	type	required	default	help
study	argument	str	yes		The study id: its directory under safety/hazop/.
--input	option	str			Add an analysed input (repeat).
--write	option	flag			Record the revisions.

cheme safety hazop review

Prepare a review record (skill review's format) for the study: scope, basis (analysed inputs and matrix) and the open items. The reviewer is left empty: the record does not validate until a person fills it in.

cheme safety hazop review [OPTIONS] {study}

name	kind	type	required	default	help
study	argument	str	yes		The study id: its directory under safety/hazop/.
--write	option	flag			Write reviews/<today>-hazop-STUDY.md (default: print it).

cheme experiments

Experiments: kinetic data, rate laws fitted to them, data sampled from a mechanism, measured electrochemical series.

cheme experiments [OPTIONS] COMMAND [ARGS]...

Subcommands: fit-rate, sample-mechanism, electrochem

cheme experiments fit-rate

Fit a rate law to kinetic data: the law as a rate: entry for process/reactions.yaml, each parameter with its standard error and 95 % interval, and the R^2 of ln r.

cheme experiments fit-rate [OPTIONS] {data}

name	kind	type	required	default	help
data	argument	path	yes		A kinetic data file (cheme/kinetic-data/v1).
--form	option	str		power_law	power_law or lhhw.
--order	option	str			Power law: a component with an order (repeat; default: the reactants). LHHW: ID=EXPONENT of the numerator.
--adsorb	option	str			LHHW: a component in the denominator (repeat).
--power	option	float		1.0	LHHW: the denominator's exponent.
--reverse	option	flag			Fit r / (1 - Q/K): the reaction is reversible.
--json	option	flag			As JSON.

cheme experiments sample-mechanism

Sample a reaction's mechanism over a grid of conditions into a kinetic data file, to fit a rate law for a fixed bed with cheme experiments fit-rate.

cheme experiments sample-mechanism [OPTIONS] {reaction}

name	kind	type	required	default	help
reaction	argument	str	yes		A reaction with a mechanism.
--temperature / -T	option	float	yes		K (repeat).
--pressure / -p	option	str	yes		ID=BAR,BAR,... partial pressures (repeat).
--sites-per-kg	option	float	yes		kmol of active sites per kg of catalyst.
--out	option	path	yes		The kinetic data file to write.
--product	option	str			Default: the first product.

cheme experiments electrochem

Process a measured electrochemical series: per replicate and point (mean, sd, n) the Faradaic efficiency of each product, partial current densities, the water dissociation efficiency from the anolyte titration, power, heat and the charge and carbon balances. Reads the file only (a plant repository is not needed); --plot writes polarisation, Faradaic efficiency, partial current density and water dissociation plots. Exit 1 on an error finding, 2 when the file does not exist.

cheme experiments electrochem [OPTIONS] {file}

name	kind	type	required	default	help
file	argument	path	yes		A measurement file (cheme/electrochemical-measurements/v1).
--json	option	flag			As JSON.
--plot	option	path			Write the SVG figures the data support into this directory.

cheme data

Optional datasets: fetch them into the user data directory.

cheme data [OPTIONS] COMMAND [ARGS]...

Subcommands: fetch, status

cheme data fetch

Download (or read with --zip) an optional dataset, check it and install it.

chemsep-srk: ChemSep-LITE from http://www.chemsep.org/downloads/data/lite.zip (plain HTTP; the sha256 recorded for ChemSep-LITE 8.50 is checked), its ipd/srk.ipd converted to srk.json for the Redlich-Kwong-Soave flash. Its licence is not confirmed for redistribution, so cheme-as-code does not ship it. Exits 2 on a failed check.

cheme data fetch [OPTIONS] {dataset}

name	kind	type	required	default	help
dataset	argument	str	yes		Dataset: chemsep-srk.
--zip	option	file			A local ChemSep-LITE lite.zip (offline); without it the archive is downloaded.
--url	option	str			Download from this URL instead.
--sha256	option	str			Expected sha256 of the archive (default: the recorded release).
--no-verify	option	flag			Do not check the archive's sha256.
--data-dir	option	path			Install under this directory (default: $CHEME_DATA_DIR or the user data directory).

cheme data status

Which optional datasets are installed, and where.

cheme data status [OPTIONS]

name	kind	type	required	default	help
--data-dir	option	path			Look under this directory instead.

cheme runs

Run records: one folder per design session in runs/ (start, record, list).

cheme runs [OPTIONS] COMMAND [ARGS]...

Subcommands: start, record, list

cheme runs start

Remember the start of a session (HEAD, branch, time, working tree) in build/run-records/<session_id>.start.json, for cheme runs record.

cheme runs start [OPTIONS]

name	kind	type	required	default	help
--hook	option	flag			Claude Code SessionStart hook: read its JSON from stdin, never fail.
--session-id	option	str			The session id (without --hook; required then).

cheme runs record

Write runs/<run-id>/ for one session: run.yaml (process-run/v1), task.md, report.md and session.jsonl.gz. Prints the run folder. A record that exists is never rewritten.

cheme runs record [OPTIONS]

name	kind	type	required	default	help
--hook	option	flag			Claude Code SessionEnd hook: JSON from stdin, kind interactive, no checks, no commit, skipped when nothing changed, never fails.
--session-id	option	str			The Claude Code session id.
--transcript	option	path			The session transcript (.jsonl) instead of --session-id.
--kind	option	str		interactive	agent, interactive or manual.
--slug	option	str			Appended to the run id (e.g. the stage id).
--task-file	option	path			The task, copied to task.md.
--report-file	option	path			The session's report, copied to report.md.
--cost	option	float			USD as reported by the API (never an estimate).
--budget	option	float			The session's USD cap.
--effort	option	str			The reasoning effort it ran with.
--result	option	str			How it ended (Claude Code's result subtype: success, ...).
--start-commit	option	str			HEAD at the start, when no start file was written.
--checks	option	flag			Run cheme check and cheme maturity and record them.
--commit	option	flag			Commit the run folder alone (git commit -- runs/ID).
--sdk-version	option	str			Claude Agent SDK version, from a runner that uses it.
--runner-name	option	str			The external runner that ran the session.
--runner-commit	option	str			The runner's commit, when it knows one.
--runner-dirty / --runner-clean	option	flag			Whether the runner's checkout had changes.

cheme runs list

One line per run record in runs/: run id, kind, cost, wall time, complete, turns. Read-only.

cheme runs list [OPTIONS]

name	kind	type	required	default	help
--json	option	flag			Machine-readable.