pks is the propstore command-line interface -- a scientific claim compiler and reasoning engine. All commands should be run via uv run pks.
uv run pks <command> [options]| Option | Type | Description |
|---|---|---|
-C, --directory |
PATH | Run as if pks was started in this directory |
Initialize a new propstore project directory. Creates the standard directory structure (concepts/, claims/, contexts/, forms/, justifications/, sidecar/, sources/, stances/, worldlines/). If no DIRECTORY argument is given, it creates knowledge/ in the current working directory.
uv run pks init
uv run pks init my-projectValidate everything, build sidecar, run conflict detection.
| Option | Type | Default | Description |
|---|---|---|---|
-o, --output |
TEXT | -- | Output path |
--force |
FLAG | false | Force rebuild |
uv run pks build
uv run pks build --force -o out/sidecar.sqliteValidate all concepts and claims. Runs CEL type-checking. No options.
uv run pks validateRun raw SQL against the sidecar SQLite.
uv run pks query "SELECT * FROM claims WHERE concept_id LIKE 'speech%'"Build sidecar from a historical commit (non-destructive).
uv run pks checkout abc1234Inspect the formal merge framework between two branches. Prints a YAML formal_merge_report with framework_type: partial_argumentation_framework, the merged argument universe, uncertain relations, and acceptance summary for the requested semantics.
| Option | Type | Default | Description |
|---|---|---|---|
--semantics |
grounded|preferred|stable |
grounded | Completion semantics used for the summary |
uv run pks merge inspect agent/paper-a agent/paper-b
uv run pks merge inspect agent/paper-a agent/paper-b --semantics preferredCreate a two-parent storage merge commit from the formal merge framework between two branches. Prints YAML storage_merge_commit metadata with the branch pair, target branch, claims/merged.yaml, merge/manifest.yaml, and the resulting commit SHA.
| Option | Type | Default | Description |
|---|---|---|---|
--message |
TEXT | -- | Override the default merge commit message |
--target-branch |
TEXT | master | Branch ref that receives the merge commit |
uv run pks merge commit agent/paper-a agent/paper-b
uv run pks merge commit agent/paper-a agent/paper-b --target-branch synthesis --message "Merge paper branches"Import the committed semantic snapshot of another propstore repo onto a destination branch. Prints YAML repo_import_commit metadata with the source repo path, source commit, target branch, commit SHA, touched semantic paths, and whether the destination worktree was synced.
| Option | Type | Default | Description |
|---|---|---|---|
--target-branch |
TEXT | import/<repo-name> |
Branch ref that receives the import commit |
--message |
TEXT | -- | Override the default import commit message |
uv run pks -C repo-a import-repo ../repo-b
uv run pks -C repo-a import-repo ../repo-b --target-branch masterShow branch-scoped knowledge repository history with operation labels.
| Option | Type | Default | Description |
|---|---|---|---|
-n, --count |
INTEGER | 20 | Number of entries to show |
--branch |
TEXT | master |
Branch history to inspect |
--show-files |
FLAG | false |
Include per-commit added/modified/deleted paths |
--format |
text|yaml |
text |
Render as text or structured YAML |
uv run pks log
uv run pks log --branch agent/paper-a
uv run pks log --show-files
uv run pks log --format yaml
uv run pks log -n 50Show details of a specific commit.
uv run pks show abc1234Show files changed between HEAD and COMMIT (or HEAD vs parent).
uv run pks diff
uv run pks diff abc1234Promote committed stance proposals from the proposal/stances branch into source-of-truth storage on master. If PATH is given, promotes only that stance file or claim-id-derived filename; otherwise promotes all committed stance proposal files on the proposal branch.
| Option | Type | Default | Description |
|---|---|---|---|
-y, --yes |
FLAG | false | Skip confirmation prompt |
uv run pks promote
uv run pks promote some-stance.yaml -yManage source branches and source-local artifacts. The source workflow replaces the old paper-import bridge.
Key subcommands:
pks source init <name> --kind <kind> --origin-type <type> --origin-value <value> [--content-file FILE]pks source write-notes <name> --file notes.mdpks source write-metadata <name> --file metadata.jsonpks source propose-concept <name> --name <concept> --definition <text> --form <form>pks source add-concepts <name> --batch concepts.yamlpks source add-claim <name> --batch claims.yamlpks source add-justification <name> --batch justifications.yamlpks source add-stance <name> --batch stances.yamlpks source finalize <name>pks source promote <name>pks source sync <name> [--output-dir DIR]
uv run pks source init Demo_2026 --kind academic_paper --origin-type file --origin-value Demo_2026.pdf
uv run pks source write-notes Demo_2026 --file ../research/papers/Demo_2026/notes.md
uv run pks source write-metadata Demo_2026 --file ../research/papers/Demo_2026/metadata.json
uv run pks source propose-concept Demo_2026 --name pitch --definition "Fundamental frequency of voiced speech" --form frequency
uv run pks source add-claim Demo_2026 --batch ../research/papers/Demo_2026/claims.yaml
uv run pks source finalize Demo_2026
uv run pks source promote Demo_2026
uv run pks source sync Demo_2026Export the alias lookup table.
| Option | Type | Default | Description |
|---|---|---|---|
--format |
text|json |
-- | Output format |
uv run pks export-aliases --format jsonAdd a new concept to the registry.
| Option | Type | Default | Description |
|---|---|---|---|
--domain |
TEXT | -- | Domain prefix (e.g. speech, a11y, finance) (required) |
--name |
TEXT | -- | Canonical name (lowercase, underscored) (required) |
--definition |
TEXT | -- | Definition (prompted if omitted) |
--form |
TEXT | -- | Form name (references forms/<name>.yaml, prompted if omitted) |
--values |
TEXT | -- | Comma-separated values (required for category concepts) |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept add --domain speech --name pitch --form frequency --definition "Fundamental frequency of voiced speech"Add a value to a category concept's value set.
| Option | Type | Default | Description |
|---|---|---|---|
--value |
TEXT | -- | Value to add (required) |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept add-value speech.vowel_quality --value "mid-central"Add an alias to a concept.
| Option | Type | Default | Description |
|---|---|---|---|
--name |
TEXT | -- | Alias name (required) |
--source |
TEXT | -- | Source paper or 'common' (required) |
--note |
TEXT | -- | Optional note |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept alias speech.pitch --name "F0" --source "common"List all category concepts and their allowed values.
| Option | Type | Default | Description |
|---|---|---|---|
--json |
FLAG | false | Output as JSON |
uv run pks concept categoriesDeprecate a concept with a replacement.
| Option | Type | Default | Description |
|---|---|---|---|
--replaced-by |
TEXT | -- | Replacement concept ID (required) |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept deprecate speech.old_name --replaced-by speech.new_nameGenerate embeddings for concepts via litellm. See Embedding options.
| Option | Type | Default | Description |
|---|---|---|---|
--all |
FLAG | false | Embed all concepts |
--model |
TEXT | -- | litellm model string, or 'all' (required) |
--batch-size |
INTEGER | -- | Concepts per API call |
uv run pks concept embed --all --model text-embedding-3-small
uv run pks concept embed speech.pitch --model allAdd a relationship between concepts.
| Argument | Required | Description |
|---|---|---|
SOURCE_ID |
yes | Source concept |
RELATION |
yes | One of: broader, narrower, related, component_of, derived_from, contested_definition |
TARGET_ID |
yes | Target concept |
| Option | Type | Default | Description |
|---|---|---|---|
--source |
TEXT | -- | Source paper |
--note |
TEXT | -- | Note |
--conditions |
TEXT | -- | Comma-separated CEL expressions |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept link speech.pitch derived_from speech.glottal_pulse_rate --source "Titze_1994"List concepts, optionally filtered.
| Option | Type | Default | Description |
|---|---|---|---|
--domain |
TEXT | -- | Filter by domain |
--status |
TEXT | -- | Filter by status |
uv run pks concept list --domain speechRename a concept (updates canonical_name and filename).
| Option | Type | Default | Description |
|---|---|---|---|
--name |
TEXT | -- | New canonical name (required) |
--dry-run |
FLAG | false | Dry-run |
uv run pks concept rename speech.old_name --name new_nameSearch concepts by name, definition, or alias.
uv run pks concept search "frequency"Show full concept YAML.
uv run pks concept show speech.pitchFind similar concepts by embedding distance. See Embedding options.
| Option | Type | Default | Description |
|---|---|---|---|
--model |
TEXT | -- | litellm model string (default: first available) |
--top-k |
INTEGER | -- | Number of results |
--agree |
FLAG | false | Similar under ALL stored models |
--disagree |
FLAG | false | Similar under some models but not others |
uv run pks concept similar speech.pitch --top-k 5Compare two algorithm claims for equivalence.
| Option | Type | Default | Description |
|---|---|---|---|
-b, --bindings |
TEXT | -- | Known values as key=value pairs |
uv run pks claim compare algo_claim_1 algo_claim_2Detect and report claim conflicts.
| Option | Type | Default | Description |
|---|---|---|---|
--concept |
TEXT | -- | Filter by concept ID |
--class |
CONFLICT|OVERLAP|PARAM_CONFLICT |
-- | Filter by warning class |
uv run pks claim conflicts
uv run pks claim conflicts --concept speech.pitch --class CONFLICTGenerate embeddings for claims via litellm. See Embedding options.
| Option | Type | Default | Description |
|---|---|---|---|
--all |
FLAG | false | Embed all claims |
--model |
TEXT | -- | litellm model string, or 'all' (required) |
--batch-size |
INTEGER | -- | Claims per API call |
uv run pks claim embed --all --model text-embedding-3-smallClassify epistemic relationships between similar claims via LLM. The resulting stance proposal snapshots are committed to the proposal/stances branch rather than written into source-of-truth state.
| Option | Type | Default | Description |
|---|---|---|---|
--all |
FLAG | false | Relate all claims |
--model |
TEXT | -- | LLM model for classification (required) |
--embedding-model |
TEXT | -- | Embedding model for similarity |
--top-k |
INTEGER | -- | Number of similar claims to classify |
--concurrency |
INTEGER | -- | Max concurrent LLM calls |
--second-pass-threshold |
FLOAT | -- | Distance threshold for second-pass NLI |
uv run pks claim relate --all --model gpt-4o --embedding-model text-embedding-3-smallDisplay details of a single claim.
uv run pks claim show speech.pitch.claim_01Find similar claims by embedding distance. See Embedding options.
| Option | Type | Default | Description |
|---|---|---|---|
--model |
TEXT | -- | litellm model string (default: first available) |
--top-k |
INTEGER | -- | Number of results |
--agree |
FLAG | false | Similar under ALL stored models |
--disagree |
FLAG | false | Similar under some models but not others |
uv run pks claim similar speech.pitch.claim_01 --top-k 10Validate all claim files.
| Option | Type | Default | Description |
|---|---|---|---|
--dir |
TEXT | -- | Claims directory |
--concepts-dir |
TEXT | -- | Concepts directory |
uv run pks claim validateValidate a single claims YAML file.
| Option | Type | Default | Description |
|---|---|---|---|
--concepts-dir |
TEXT | -- | Concepts directory |
uv run pks claim validate-file knowledge/claims/speech/pitch.yamlAdd a new form definition.
| Option | Type | Default | Description |
|---|---|---|---|
--name |
TEXT | -- | Form name (lowercase, underscored) (required) |
--unit |
TEXT | -- | Primary unit symbol (e.g. Hz, Pa) |
--qudt |
TEXT | -- | QUDT IRI (e.g. qudt:HZ) |
--base |
TEXT | -- | Base type (e.g. ratio) |
--dimensions |
TEXT | -- | JSON dict of SI dimension exponents, e.g. '{"T": -1}' |
--dimensionless |
TEXT | -- | Whether the form is dimensionless (true/false) |
--common-alternatives |
TEXT | -- | JSON array of alternative unit conversions |
--note |
TEXT | -- | Human-readable note about this form |
--dry-run |
FLAG | false | Dry-run |
uv run pks form add --name frequency --unit Hz --qudt qudt:HZ --dimensions '{"T": -1}'List all available forms.
| Option | Type | Default | Description |
|---|---|---|---|
--dims |
TEXT | -- | Filter by dimensions (e.g. M:1,L:1,T:-2). Implies showing dims column. |
--show-dims |
FLAG | false | Show dimensions column |
uv run pks form list
uv run pks form list --dims "T:-1"Remove a form definition.
| Option | Type | Default | Description |
|---|---|---|---|
--force |
FLAG | false | Remove even if concepts reference this form |
--dry-run |
FLAG | false | Dry-run |
uv run pks form remove old_form --forceShow full form definition YAML, plus algebra if sidecar exists.
uv run pks form show frequencyValidate form definitions (one or all). Checks that every form YAML has a valid name field and that forms referenced by concepts actually exist.
uv run pks form validate
uv run pks form validate frequencyAdd a new context to the registry.
| Option | Type | Default | Description |
|---|---|---|---|
--name |
TEXT | -- | Context ID (e.g., ctx_atms_tradition) (required) |
--description |
TEXT | -- | Short description (required) |
--inherits |
TEXT | -- | Parent context ID |
--excludes |
TEXT | -- | Comma-separated excluded context IDs |
--dry-run |
FLAG | false | Dry-run |
uv run pks context add --name ctx_atms_tradition --description "Papers from the ATMS research tradition"List all registered contexts. No options.
uv run pks context listMany world commands accept positional KEY=VALUE bindings to scope queries by condition. These are passed as trailing arguments.
Show knowledge base stats (concepts, claims, conflicts). No options.
uv run pks world statusShow all claims for a concept.
uv run pks world query speech.pitchShow active claims under condition bindings. Arguments with = are bindings; the last argument without = is a concept filter.
uv run pks world bind domain=speech
uv run pks world bind domain=speech speech.pitchDerive a value for a concept via parameterization relationships.
uv run pks world derive speech.formant_ratio domain=speechTraverse the parameter space to derive a target concept.
| Option | Type | Default | Description |
|---|---|---|---|
--strategy |
recency|sample_size|argumentation|override |
-- | Resolution strategy |
uv run pks world chain speech.formant_ratio domain=speech --strategy sample_sizeResolve a conflicted concept using a strategy. Supports Argumentation options, PrAF options, and Decision criterion options.
| Option | Type | Default | Description |
|---|---|---|---|
--strategy |
recency|sample_size|argumentation|override |
-- | Resolution strategy (required) |
--override |
TEXT | -- | Claim ID for override strategy |
uv run pks world resolve speech.pitch domain=speech --strategy argumentation --semantics preferredShow the stance chain for a claim.
uv run pks world explain speech.pitch.claim_01Show what changes if claims are removed or added.
| Option | Type | Default | Description |
|---|---|---|---|
--remove |
TEXT | -- | Claim ID to remove |
--add |
TEXT | -- | JSON synthetic claim |
uv run pks world hypothetical domain=speech --remove speech.pitch.claim_02Show the current finite revision belief base for a scoped world.
uv run pks world revision-base speaker_sex=maleShow the current deterministic entrenchment ordering for that scoped revision base.
uv run pks world revision-entrenchment speaker_sex=maleRun one-shot expansion over the derived revision base. Does not mutate source YAML.
| Option | Type | Default | Description |
|---|---|---|---|
--atom |
TEXT | -- | JSON revision atom to add (required) |
uv run pks world expand speaker_sex=male --atom '{"kind":"claim","id":"synthetic_freq","value":123.0}'Run one-shot contraction over the derived revision base. Does not mutate source YAML.
| Option | Type | Default | Description |
|---|---|---|---|
--target |
TEXT | -- | Existing atom or claim id to contract (repeatable, required) |
uv run pks world contract speaker_sex=male --target freq_claim1Run one-shot revision over the derived revision base. Does not mutate source YAML.
| Option | Type | Default | Description |
|---|---|---|---|
--atom |
TEXT | -- | JSON revision atom to admit (required) |
--conflict |
TEXT | -- | Existing atom or claim id that conflicts with the new atom |
uv run pks world revise speaker_sex=male --atom '{"kind":"claim","id":"synthetic_freq","value":123.0}' --conflict freq_claim1Explain one one-shot revision operation over the current scoped world.
| Option | Type | Default | Description |
|---|---|---|---|
--operation |
expand|contract|revise |
-- | Revision operation to explain (required) |
--atom |
TEXT | -- | JSON revision atom for expand / revise |
--target |
TEXT | -- | Existing atom or claim id for contract |
--conflict |
TEXT | -- | Existing atom or claim id that conflicts with the new atom |
uv run pks world revision-explain speaker_sex=male --operation contract --target freq_claim1Inspect the current explicit iterated revision state for a scoped world.
uv run pks world iterated-state speaker_sex=maleRun one iterated revision episode and print the next explicit epistemic state. This remains a revision-state transform, not a worldline.
| Option | Type | Default | Description |
|---|---|---|---|
--atom |
TEXT | -- | JSON revision atom to admit (required) |
--conflict |
TEXT | -- | Existing atom or claim id that conflicts with the new atom |
--operator |
restrained|lexicographic |
restrained | Iterated operator family |
uv run pks world iterated-revise speaker_sex=male --atom '{"kind":"claim","id":"synthetic_freq","value":123.0}' --conflict freq_claim1 --operator lexicographicAnalyze which input most influences a derived quantity.
| Option | Type | Default | Description |
|---|---|---|---|
--format |
text|json |
-- | Output format |
uv run pks world sensitivity speech.formant_ratio domain=speechExport the knowledge graph as DOT or JSON.
| Option | Type | Default | Description |
|---|---|---|---|
--format |
dot|json |
-- | Output format |
--group |
INTEGER | -- | Parameterization group ID to filter by |
--output |
TEXT | -- | Output file path |
uv run pks world export-graph domain=speech --format dot --output graph.dotCheck for conflicts, optionally including transitive (multi-hop) ones.
| Option | Type | Default | Description |
|---|---|---|---|
--transitive |
FLAG | false | Check multi-hop transitive conflicts |
uv run pks world check-consistency domain=speech --transitiveList algorithm claims in the world model.
| Option | Type | Default | Description |
|---|---|---|---|
--stage |
TEXT | -- | Filter by processing stage |
--concept |
TEXT | -- | Filter by concept |
uv run pks world algorithms --stage preprocessingRank epistemic targets by fragility -- what to learn next.
| Option | Type | Default | Description |
|---|---|---|---|
--concept |
TEXT | -- | Focus on a single concept |
--top-k |
INTEGER | -- | Number of results |
--combination |
top2|mean|max|product |
-- | Combination method |
--skip-parametric |
FLAG | false | Skip parametric fragility |
--skip-epistemic |
FLAG | false | Skip epistemic fragility |
--skip-conflict |
FLAG | false | Skip conflict fragility |
--sort-by |
fragility|roi |
-- | Sort order |
--discovery-tier |
INTEGER | -- | 1=ATMS only, 2=also unknown concepts |
--format |
text|json |
-- | Output format |
uv run pks world fragility --top-k 10 --sort-by roiShow argumentation extensions -- all claims that survive scrutiny. Supports Argumentation options and PrAF options.
| Option | Type | Default | Description |
|---|---|---|---|
--backend |
claim_graph|aspic|atms|praf |
claim_graph | Argumentation backend |
uv run pks world extensions domain=speech --semantics grounded
uv run pks world extensions domain=speech --backend praf --praf-strategy mcAll ATMS commands support ATMS options.
Show ATMS-native claim status, support quality, and essential support.
uv run pks world atms-status --context ctx_atms_traditionShow which ATMS-supported claims hold in the current bound environment.
uv run pks world atms-context --context ctx_atms_traditionRun ATMS label self-checks for the current bound environment.
uv run pks world atms-verifyShow bounded ATMS future environments for a claim or concept.
| Option | Type | Default | Description |
|---|---|---|---|
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) (required) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-futures speech.pitch --queryable "domain=speech" --limit 4Show bounded ATMS-native stability over the implemented future replay substrate.
| Option | Type | Default | Description |
|---|---|---|---|
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) (required) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-stability speech.pitch --queryable "domain=speech"Show which bounded queryables can flip an ATMS or concept status.
| Option | Type | Default | Description |
|---|---|---|---|
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) (required) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-relevance speech.pitch --queryable "domain=speech"Show bounded additive intervention plans for an ATMS claim or concept.
| Option | Type | Default | Description |
|---|---|---|---|
--target-status |
TEXT | -- | Desired ATMS node status (IN/OUT) (required) |
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) (required) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-interventions speech.pitch --target-status IN --queryable "domain=speech"Show next-query suggestions derived from bounded additive intervention plans.
| Option | Type | Default | Description |
|---|---|---|---|
--target-status |
TEXT | -- | Desired ATMS node status (IN/OUT) (required) |
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) (required) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-next-query speech.pitch --target-status IN --queryable "domain=speech"Explain whether an ATMS OUT status is missing-support or nogood-pruned.
| Option | Type | Default | Description |
|---|---|---|---|
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
uv run pks world atms-why-out speech.pitch --queryable "domain=speech"Materialized query artifacts -- traced paths through the knowledge space with full provenance.
Create a worldline definition (question only, no results yet). Supports Argumentation options, PrAF options, Decision criterion options, ASPIC+ options, and Revision Worldline Options.
| Option | Type | Default | Description |
|---|---|---|---|
--bind |
TEXT | -- | Condition binding (key=value) |
--with |
TEXT | -- | Value override (concept=value) |
--target |
TEXT | -- | Target concept to derive/resolve (repeatable, required) |
--strategy |
recency|sample_size|argumentation|override |
-- | Resolution strategy |
--context |
TEXT | -- | Context to scope the query |
--revision-operation |
expand|contract|revise|iterated_revise |
-- | Optional revision operation to attach |
--revision-atom |
TEXT | -- | JSON revision atom for expand / revise / iterated_revise |
--revision-target |
TEXT | -- | Revision target for contract |
--revision-conflict |
TEXT | -- | Conflict mapping atom_id=target[,target...]; repeatable |
--revision-operator |
restrained|lexicographic |
-- | Iterated revision operator family |
uv run pks worldline create my-worldline --target speech.pitch --bind domain=speech --strategy argumentationRun (materialize) a worldline. Creates it first if it doesn't exist. Supports Argumentation options, PrAF options, Decision criterion options, ASPIC+ options, and Revision Worldline Options.
| Option | Type | Default | Description |
|---|---|---|---|
--bind |
TEXT | -- | Condition binding (key=value) |
--with |
TEXT | -- | Value override (concept=value) |
--target |
TEXT | -- | Target concept (repeatable; required only when creating a new worldline) |
--strategy |
recency|sample_size|argumentation|override |
-- | Resolution strategy |
--context |
TEXT | -- | Context scope |
--revision-operation |
expand|contract|revise|iterated_revise |
-- | Optional revision operation to attach |
--revision-atom |
TEXT | -- | JSON revision atom for expand / revise / iterated_revise |
--revision-target |
TEXT | -- | Revision target for contract |
--revision-conflict |
TEXT | -- | Conflict mapping atom_id=target[,target...]; repeatable |
--revision-operator |
restrained|lexicographic |
-- | Iterated revision operator family |
uv run pks worldline run my-worldlineShow a worldline's results, including any stored revision query and revision result payload.
| Option | Type | Default | Description |
|---|---|---|---|
--check |
FLAG | false | Check for staleness |
uv run pks worldline show my-worldline --checkList all worldlines. No options.
uv run pks worldline listCompare two worldlines side by side.
uv run pks worldline diff baseline experimentalRe-run a worldline with current knowledge.
uv run pks worldline refresh my-worldlineDelete a worldline.
uv run pks worldline delete old-worldlineSeveral option sets recur across multiple commands. They are defined here once and referenced by name from individual commands.
Used by: world extensions, world resolve, worldline create, worldline run.
| Option | Type | Default | Description |
|---|---|---|---|
--reasoning-backend |
claim_graph|aspic|atms|praf |
claim_graph | Argumentation backend |
--semantics |
backend-dependent | grounded | Argumentation semantics. worldline create/run accepts grounded, legacy_grounded, hybrid-grounded, bipolar-grounded, preferred, stable, d-preferred, s-preferred, c-preferred, bipolar-stable, and complete; backend support is validated separately. |
--set-comparison |
elitist|democratic |
elitist | Set comparison for preference ordering |
--context |
TEXT | -- | Context to scope the argumentation |
Note: world extensions uses --backend instead of --reasoning-backend.
Note: world resolve and world extensions currently accept only grounded, preferred, and stable for --semantics.
structured_projection remains accepted only as a legacy alias in parsed policy inputs, not as a CLI option value.
Used alongside argumentation options when --backend/--reasoning-backend is praf.
| Option | Type | Default | Description |
|---|---|---|---|
--praf-strategy |
auto|mc|exact|dfquad_quad|dfquad_baf |
auto | PrAF computation strategy |
--praf-epsilon |
FLOAT | 0.01 | PrAF MC error tolerance |
--praf-confidence |
FLOAT | 0.95 | PrAF MC confidence level |
--praf-seed |
INTEGER | random | PrAF MC RNG seed |
Used by: world resolve, worldline create, worldline run.
| Option | Type | Default | Description |
|---|---|---|---|
--decision-criterion |
pignistic|lower_bound|upper_bound|hurwicz |
pignistic | Decision criterion for opinion interpretation |
--pessimism-index |
FLOAT | 0.5 | Hurwicz pessimism index (0-1) |
Used by: worldline create, worldline run.
| Option | Type | Default | Description |
|---|---|---|---|
--link-principle |
last|weakest |
last | ASPIC+ link principle (Modgil & Prakken 2018, Defs 19-21) |
Used by: worldline create, worldline run.
| Option | Type | Default | Description |
|---|---|---|---|
--revision-operation |
expand|contract|revise|iterated_revise |
-- | Optional revision operation attached to the worldline |
--revision-atom |
TEXT | -- | JSON revision atom for expand / revise / iterated_revise |
--revision-target |
TEXT | -- | Revision target for contract |
--revision-conflict |
TEXT | -- | Conflict mapping atom_id=target[,target...]; repeatable |
--revision-operator |
restrained|lexicographic |
-- | Iterated revision operator family |
Used by all world atms-* commands.
| Option | Type | Default | Description |
|---|---|---|---|
--context |
TEXT | -- | Context to scope the ATMS inspection |
Commands that support future-environment exploration (atms-futures, atms-stability, atms-relevance, atms-interventions, atms-next-query, atms-why-out) additionally accept:
| Option | Type | Default | Description |
|---|---|---|---|
--queryable |
TEXT | -- | Future/queryable assumption (CEL or key=value) |
--limit |
INTEGER | 8 | Maximum number of future environments to inspect |
Used by: concept embed, concept similar, claim embed, claim similar.
| Option | Type | Default | Description |
|---|---|---|---|
--model |
TEXT | -- | litellm model string (or 'all' for embed commands) |
--top-k |
INTEGER | -- | Number of results (similar commands) |
--agree |
FLAG | false | Similar under ALL stored models (similar commands) |
--disagree |
FLAG | false | Similar under some models but not others (similar commands) |
Used by: concept add, concept add-value, concept alias, concept deprecate, concept link, concept rename, context add, form add, form remove.
| Option | Type | Default | Description |
|---|---|---|---|
--dry-run |
FLAG | false | Show what would happen without writing |
Used with pks world resolve --strategy and pks worldline create/run --strategy.
| Strategy | Behavior |
|---|---|
recency |
Most recent paper wins |
sample_size |
Largest sample size wins |
argumentation |
Run argumentation backend, project survivors |
override |
Specify a claim ID directly via --override |