CLI Reference
Complete reference for all amc commands — organized by namespace with examples, flags, and environment variables.
By the numbers
The AMC CLI currently registers 1,171 command paths. Every command supports --help for inline documentation.
Common Workflow Commands
The commands you'll use most, in order of typical workflow:
Command What it does Example
amcDetect setup and run the full maturity score amcamc setup --demoCreate demo data and run the same full-score path amc setup --demo && amcamc doctorCheck runtime and workspace health amc doctoramc fixScore, explain the gaps, and write guardrails into your agent's config amc fix --target-level L4amc guide --goAuto-detect framework and apply guardrails amc guide --goamc upStart Studio (gateway + dashboard) amc upamc statusShow Studio and vault status amc statusamc evidence collectInteractive evidence collection wizard amc evidence collectamc wrapWrap any agent CLI for evidence capture amc wrap claude -- claude "hello"amc adapters runRun adapter with lease + gateway routing amc adapters run --agent my-agent --adapter claude-cli -- claudeamc adapters capabilitiesIssue a signed declared/effective capability receipt amc adapters capabilities --agent my-agent --adapter claude-cli --jsonamc policy controlsInspect verified Scope, When, Then, and Status across existing signed controls amc policy controls --jsonamc policy simulatePreview one control through its production evaluator without recording amc policy simulate action:DEPLOY --risk high --mode executeamc policy testRun deterministic policy fixtures through production evaluators amc policy test policy-fixtures.yaml --jsonamc runRun full maturity diagnostic amc run --agent my-agent --window 30amc run --question-set lifecycleRun the explicit 264-question lifecycle-expanded diagnostic amc run --question-set lifecycleamc proof checkCheck a claim against a declared source-to-rule manifest and emit an amcproof artifact amc proof check --domain governance --manifest fixtures/domain-proof/toy-governance/source-rule-manifest.json --input examples/domain-proof/toy-governance/proven.jsonamc evidence finding-proofsTrace findings to evidence, resources, receipts, and recommendations amc evidence finding-proofs listamc reportRender diagnostic report by run ID, alias, prefix, or latest amc report q1-assessment --share --public-base-url https://reports.example.com/amcamc badgeGenerate maturity badge amc badge --format htmlamc assurance runRun red team assurance packs amc assurance run --scope fullamc compliance reportGenerate compliance mapping report amc compliance report --framework EU_AI_ACTamc domain assessAssess against an industry domain amc domain assess --domain healthamc vault initInitialize encrypted key vault amc vault initamc fleet overviewExecutive fleet summary with verdict, coverage, drift, and next actions amc fleet overview --jsonamc fleet trust-graphExport delegation trust topology as Mermaid, DOT, or JSON amc fleet trust-graph --format mermaidamc compareCompare two diagnostic runs amc compare <runA> <runB>amc downStop Studio amc down
Search Commands
All Commands by Namespace
Expand All
Setup & Health 7 commands
Command Description
amc initInitialize .amc workspace amc setup [--demo]Deterministic go-live setup (--demo for sample data) amc bootstrapBootstrap workspace for production (non-interactive) amc doctor [--strict]Check install or workspace readiness; --strict requires an initialized healthy workspace amc doctor-fixAuto-repair common setup issues amc quickstartOne-command setup wizard (legacy) amc fix-signaturesVerify and re-sign gateway/fleet/agent configs
Studio & Control Plane 9 commands
Command Description
amc upStart Studio local control plane with lifecycle evidence, gateway, proxy, dashboard, and API amc downStop Studio amc statusShow Studio and vault status amc logsPrint latest Studio logs amc studio pingPing Studio API /health endpoint amc studio startStart Studio in foreground (deployment-safe) amc studio healthcheckHealth/readiness probe for deployment amc studio lan enableEnable LAN mode with pairing gate amc studio lan disableDisable LAN mode, revert to localhost
Configuration 2 commands
Command Description
amc config printPrint resolved runtime config (secret-safe) amc config explainExplain config source precedence and risky settings
Adapters 9 commands
Command Description
amc adapters initCreate signed adapters.yaml defaults amc adapters verifyVerify adapters.yaml signature amc adapters listList built-in adapters and per-agent preferences amc adapters detectDetect installed adapter runtimes amc adapters capabilitiesIssue a signed receipt for declared/effective events, controls, version proof, and lossiness amc adapters configureSet adapter profile for an agent amc adapters envPrint adapter-compatible environment exports amc adapters init-projectGenerate runnable local adapter sample amc adapters runRun adapter with minted lease, gateway routing, evidence capture
See the full Adapter Guides for framework-specific setup.
Agent Wrapping & Evidence Capture 11 commands
Command Description
amc wrap <name> -- <args>Wrap and capture evidence (claude, gemini, openclaw, any) amc supervise --route <r> -- <cmd>Supervise with gateway routing amc sandbox run --agent <id> -- <cmd>Run in hardened Docker sandbox amc monitor --stdin --runtime <n>Record stdin stream as evidence amc connectConnect wizard for any agent/provider amc connect hooks install --provider <id> [--mode observe|control]Install a reversible observer by default, or opt into loopback signed control amc connect hooks status --provider <id>Verify provider config ownership, mode, signed manifest, and narrow lease amc connect hooks health --provider <id>Join signed setup state with the latest receipt-verified provider event; encrypted evidence fails closed while Vault is locked amc connect hooks lifecycle --agent <id> --action <id>Verify requested, decision, and completed or failed receipts for one action amc connect hooks remove --provider <id>Remove only AMC's handler and revoke its observation lease amc ingestIngest external logs as SELF_REPORTED evidence amc attestAuditor-attest to upgrade trust to ATTESTED
Rapid Checks, Badge & Telemetry 8 commands
Command Description
amc quickscoreOptional rapid pulse check; use amc for the full score amc quickscore --rapidRun the lightweight pulse-check mode explicitly amc quickscore --jsonMachine-readable JSON output amc compliance report --framework EU_AI_ACTGenerate EU AI Act mapping from evidence amc badgeGenerate maturity badge (markdown, HTML, or URL) amc badge --format htmlHTML <img> tag output amc telemetry on|offEnable/disable anonymous opt-in telemetry amc telemetry statusShow telemetry config
Scoring & Diagnostics 19 commands
Command Description
amc run --agent <id> --window <d>Run maturity diagnostic amc run --question-set lifecycleOpt into lifecycle, evidence, runtime, proof, memory, and fleet question coverage amc proof check --manifest <path> --input <path>Emit a fail-closed amcproof artifact for declared source-to-rule correctness checks amc report <runId|alias|latest>Render report for a run, saved alias, or latest run amc report <id> --html r.htmlStyled HTML report (print to PDF) amc report <alias> --share --public-base-url <url>Static report share bundle with local and public URL manifest amc report <id> --executiveBoard-friendly terminal summary amc run-alias set <alias> <run>Name a diagnostic run for later reports amc run-alias listList saved run aliases for the active agent amc run-alias remove <alias>Remove a saved run alias amc historyList diagnostic run history with aliases amc compare <a> <b>Compare two runs amc observe timelineTimeline of scores, evidence volume, and anomalies amc trace inspectInspect evidence events, tool calls, and trust tiers amc dataset run <name>Run a golden dataset against a model endpoint amc lite-scoreScore a non-agent LLM app or chatbot amc business kpiCorrelate maturity with business risk and audit readiness amc verifyVerify integrity across AMC artifacts amc verify all --jsonFull verification in one pass amc snapshot --agent <id>Generate Unified Clarity Snapshot amc target setInteractive equalizer wizard amc target diffDiff run against target profile amc whatif targetsWhat-if simulation on targets amc indices --agent <id>Compute failure-risk indices amc diagnostic renderContextualized question view
Gateway, Proxy & Runtime 21 commands
Command Description
amc gateway init --provider <n>Initialize gateway config amc gateway start --config <f>Start the gateway amc gateway statusShow gateway status amc gateway verify-configVerify gateway config amc gateway bind-agentBind agent to gateway route amc firewall enable --mode observe|warn|blockRoll out one signed Runtime Firewall policy as non-blocking observation, warnings, or full enforcement amc firewall statusShow exact-policy verified would-warn/would-block, actual-action, suppression, legacy, and integrity counters amc firewall check --direction request --text "..."Preview an allow, warn, or block decision with reasons and receipt refs amc firewall eventsList signed runtime decision events amc firewall export --format splunk --redactedExport redacted SIEM-friendly decision logs amc firewall disableDisable Runtime Firewall policy for the workspace amc firewall migrate-signature --approve-legacy-kindVerify and preserve an exact legacy policy in the monotonic journal amc policy controls [--json]Read the verified Runtime Firewall, Guardrails, Action Policy, and Approval Policy projection; invalid evidence reports the real fail-closed outcome and exits 2 amc policy simulate <controlId> [--json]Return exact matched rules, structured conditions, and the safe outcome from the owning production evaluator; inputs stay transient and the result is not proof amc guardrails listCompare signed requested state with effective, verified runtime bindings amc guardrails enable <id>Add a signed request for one of three Runtime Firewall-bound controls amc guardrails disable <id>Remove the additive request without weakening signed policy amc guardrails profile <name>Apply the bound subset and disclose catalog-only exclusions amc runtime statusShow persisted connected-agent run state counts amc runtime create --run <id>Create a resumable runtime run amc runtime event <id> --type trace.receivedAppend a redacted runtime event amc runtime inspect <id>Inspect run state and event stream amc runtime resume|cancel|degrade|complete <id>Manage run lifecycle state cleanly amc runtime export <id> --redactedExport runtime run events as JSON or JSONL
Fleet & Agents 31 commands
Command Description
amc fleet initCreate and sign fleet.yaml amc fleet report --window <d>Cross-agent fleet report amc fleet overviewOne-shot executive fleet summary with verdict, coverage, drift, and next actions amc fleet graph write --file graph.jsonWrite typed nodes, handoffs, tools, contracts, policies, permissions, and invariants amc fleet graph showInspect the latest typed graph, digest, node/edge counts, and validation status amc fleet graph validateFail invalid graph contracts, unsafe permissions, cycles, and unbounded fan-out with actionable errors amc fleet graph listList saved typed multi-agent graph versions amc fleet score --all --stream --sla 120sFull-score every configured agent with progressive SLA status and parent/child lifecycle evidence amc fleet lifecycle listList fleet parent lifecycle artifacts amc fleet lifecycle show <run>Inspect child runs, topology, typed graph digest, shared resources, and cascade failures amc org rolesList the canonical 70-role operating model amc org run --roles REV_PRODUCT_MANAGER,REV_TECH_LEAD,REV_QA_LEADAdvanced Fleet role loop with isolated workspaces, Watch heartbeats, Enforce gates, and Vault-signed evidence amc org runs|inspect <run> --redactedReview org-run status, handoffs, blocked gates, receipts, and rollback refs without private grader paths amc fleet trust-initInitialize trust composition amc fleet trust-add-edgeAdd delegation edge amc fleet trust-reportTrust composition report amc fleet trust-graph --format mermaid|dot|jsonExport delegation trust graph for review, docs, or audit packets amc fleet slo define|list|statusDefine and check fleet governance SLOs such as L3+ coverage thresholds amc fleet dagVisualize delegation graph amc fleet contradictionsDetect cross-agent contradictions amc agent addAdd agent to fleet amc agent listList fleet agents amc agent removeRemove agent amc agent use <id>Set current agent amc agent diagnoseLease-auth self-run diagnostic amc lease issue --agent <id>Issue scoped lease amc lease verify <token>Verify a lease token amc lease revoke <id>Revoke a lease amc provider listList provider templates amc provider addAssign/update provider for an agent amc fleet handoffManage handoff packets
Observability & Trace 10+ commands
Command Description
amc observe timelineShow maturity progression, evidence volume, and score history amc observe anomaliesDetect volatility, regressions, and suspicious score movements amc observe compareCompare multiple agents or time windows amc correction logTrack human corrections and overrides amc correction reportSummarize correction closure and effectiveness amc trace listList recent agent sessions with evidence summary amc trace inspectInspect tool calls, decisions, payloads, and trust tiers amc evidence lifecycle list|inspect|export --redactedList, inspect, and export shareable lifecycle artifacts amc evidence episodes list|inspect|export --redactedList, inspect, and export durable full-score evidence episodes amc evidence decisions list|inspectInspect recommendation and evidence-request receipts amc evidence decisions observe <run>Update proposed decision receipts with later observed outcomes amc evidence observability list|inspectInspect component attribution, experience signals, and decision-chain records amc memory writeback <episode>Write redacted, evidence-backed reasoning memory with expiry and receipts amc memory retrieve --consumer fixerRetrieve active reasoning memory with citations for score, recommendations, fixer, or Studio amc report <run>Review confidence controls, uncertainty language, and auto-fix review gates amc trace index --run <run>Inspect distilled trace failure index entries and clusters amc trace failuresShow ranked recurring failure clusters and repair inputs amc mechanic rca run <run>Generate RCA, regression tests, rollback pointers, and governed Enforce fix proposals amc experiment optimize --rca latestCreate governed optimizer candidates with held-out validation and leakage checks amc evidence finding-proofs list|inspect|export --redactedTrace findings to evidence, resources, receipts, and recommendations amc evidence lifecycle-receipts list|inspect|export --redactedInspect proposal, validation, commit, rollback, and monitor receipts amc trace statsBreak down events by type, trust tier, and tool usage amc alert configConfigure webhook, Slack, or PagerDuty destinations amc alert send|test|watchSend manual alerts or auto-watch for anomalies
Datasets & Lite Scoring 9+ commands
Command Description
amc dataset create <name>Create a golden evaluation dataset amc dataset add-case <name>Add a prompt/assertion pair to a dataset amc dataset listList evaluation datasets amc dataset run <name>Run a dataset against your configured endpoint or model amc dataset import <name> --file <path>Import CSV/JSON/JSONL test cases amc import <path> --dry-runDetect neutral traces, event logs, run directories, graphs, configs, memory, evals, and benchmarks amc import <path>Persist redacted imported evidence into episodes, lifecycle runs, resource manifests, and trace indexes amc imports list|show|rollbackInspect and roll back neutral import runs amc strategy compare --file strategies.jsonRank inference strategies with score, cost, latency, risk, confidence, and evidence refs amc strategy compare --file strategies.json --apply --approveCommit a recommended route through policy approval, resource manifest evidence, and rollback data amc strategy list|show|rollbackInspect strategy receipts and roll back accepted route changes amc lite-scoreLite scoring for non-agent chatbots and vanilla LLM apps amc eval registry [--refresh] [--json]Inspect or explicitly refresh the signed metadata catalog for existing AMC evaluators amc eval importImport results from LangSmith, DeepEval, Promptfoo, and more amc eval statusCheck imported eval coverage
Business & Leaderboards 10+ commands
Command Description
amc business kpiMap maturity to risk reduction, compliance readiness, and audit posture amc business trackRecord incidents, findings, savings, and outcome events amc business reportGenerate business impact report with maturity correlation amc leaderboard showShow fleet-wide maturity rankings amc leaderboard exportExport leaderboard as JSON, Markdown, or HTML amc inventory scanDiscover AI agents, frameworks, models, and API-key footprints amc inventory scan --deepSearch deeper for local model artifacts and configs amc comms-check --text "..."Check messages against lightweight compliance rules
Governor & Policy 12 commands
Command Description
amc governor check --agent <id>Check if action is allowed amc governor explainExplain enforcement classification amc governor reportGenerate governance report amc policy action init|verifySigned autonomy action policy amc policy approval init|verifySigned dual-control approval policy amc policy pack list|describe|applyPolicy packs by archetype/risk amc resource snapshot|validate|apply|rollbackOne governed Enforce protocol for prompts, tools, policies, memory, routes, evaluators, datasets, schemas, environments, and configs amc resource propose|evaluate|history|contractDry-run resource changes, inspect gates, trace signed receipts, and view the AMC-native resource contract amc import <path>Map existing neutral run artifacts into AMC evidence without platform-specific connectors or public source names amc strategy compareExplain model route tradeoffs and gate route changes through receipts amc fleet graph write|validateRegister and validate typed multi-agent graph topology before fleet scoring amc fleet score --all --stream --sla 120sFull-score every configured agent with progressive status, per-agent SLA timing, lifecycle artifacts, and partial-failure summaries npm run release:gateRun the release gate and write a JSON receipt for CLI, Studio assets, OpenAPI, docs drift, domain packs, and optional live health amc commands --markdownGenerate the live CLI inventory from the registered command map so docs can stay in sync amc enforce resources snapshot|diff|verifyAdvanced aliases for the same signed Enforce resource engine amc approvals list --agent <id> [--query <requestId>] [--status <status>] [--action-class <class>] [--risk-tier <tier>] [--effective-mode <mode>] [--created-after <time>] [--created-before <time>] [--order newest|oldest] [--limit 1-200] [--json]Search the privacy-safe signed approval view; untrusted activity returns no rows amc approvals approve <id>Approve execution intent amc approvals deny <id>Deny execution intent amc workorder create|list|showSigned work order lifecycle amc archetype list|describe|applyBuilt-in role packs amc tools init|verify|list [--json]Verify signed ToolHub context and group native/MCP tools by stable server identity
Vault & Trust 10 commands
Command Description
amc vault initInitialize encrypted key vault amc vault statusShow vault status amc vault unlock|lockUnlock/lock vault amc vault rotate-keysRotate encryption keys amc notary initInitialize notary signing boundary amc notary startStart notary process amc notary attest --out <f>Generate attestation amc notary verify-attest <f>Verify attestation amc trust enable-notaryEnable notary-backed trust amc trust statusShow trust posture
Assurance Lab 10 commands
Command Description
amc assurance initInitialize assurance policy amc assurance run --scope <s>Run assurance packs (full, pack-specific) amc assurance run --verboseFull scenario detail with payloads amc assurance run --format sarifSARIF 2.1.0 export for security tools amc assurance runsList past runs amc assurance show <runId>Show run details amc shield confirm scope-write --file security-scope.jsonWrite a signed authorization scope for controlled exploit confirmation amc shield confirm run --scope <scope> --task finding-task.jsonRun fail-closed confirmation and emit safe proof without exploit instructions amc shield confirm proofs|export <proof>List and export redacted confirmation proof for Vault or Passport evidence amc assurance cert issue|verifyAssurance certificates amc assurance scheduler statusScheduler controls amc assurance waiver requestTemporary waiver amc assurance verify-policyVerify policy signature
Compliance & Audit 16 commands
Command Description
amc compliance init|verifyInitialize compliance mapping amc compliance report --framework <f>Generate compliance report amc compliance fleetFleet-wide compliance overview amc compliance diffDiff compliance between runs amc domain listList all 7 domains + packs amc domain assess --domain <d>Assess against an industry domain amc domain gaps --domain <d>Find compliance gaps amc domain report --domain <d>Generate domain report amc audit init|verify-policyInitialize audit binder amc audit binder create|verifyBinder artifact lifecycle amc audit binder create --framework eu-ai-actExport compliance evidence binder amc audit map list|show|applyCompliance map operations amc audit request create|approveAuditor evidence requests amc certificationIssue offline certificate amc cert verify|inspect|revokeCertificate operations amc passport init|create|verifyAgent Passport credential
Operations 20 commands
Command Description
amc guide --statusOne-line health check amc guide --interactiveCherry-pick which gaps to fix amc guide --watchContinuous monitoring + auto-update amc guide --diffWhat improved since last run amc guide --ci --target 3CI gate: fail build if below L3 amc fix --dry-runPreview what would be generated amc budgets init|verify|statusUsage budget management amc drift check|reportDrift/regression detection amc freeze status|liftExecution freeze controls amc alerts init|verify|testAlert configuration amc backup create|verify|restoreEncrypted backup/restore amc bundle export|verifyPortable evidence bundles amc gate --bundle <f>Release gate check amc ci init|printCI/CD release gate helpers amc loop init|run|planContinuous maturity loop amc dashboard build|serveBuild/serve local dashboard amc transparency init|verify|tailTransparency log amc transparency merkle rebuild|rootMerkle proofs amc metrics statusPrometheus metrics endpoint amc maintenance stats|vacuumOperational maintenance
Advanced 40+ commands
Command Description
amc mechanic initInitialize mechanic workspace amc mechanic gapShow maturity gaps amc mechanic plan create|executeUpgrade plan lifecycle amc mechanic rca list|showReview signed Fixer RCA reports amc learn --agent <id>Education flow for a question amc own --agent <id>Ownership flow for top gaps amc commit --agent <id>Commitment plan amc federate init|peer addCross-org federation amc outcomes init|reportOutcome contracts amc bom generate|sign|verifyMaturity Bill of Materials amc forecast init|refreshEvidence-gated forecasting amc experiment create|run|analyzeExperiment lifecycle amc experiment optimize|optimizer-list|optimizer-showGoverned optimizer candidates with held-out validation and receipts amc memory writeback|retrieve|showGoverned reasoning memory from full-score episodes amc casebook init|add|listSigned casebook operations amc plugin keygen|pack|installPlugin development & lifecycle amc release pack|verify|sbomRelease engineering amc host init|bootstrapMulti-workspace host mode amc identity initEnterprise identity config amc sso configure oidc|samlDiscoverable SSO provider setup amc scim initEnable SCIM provisioning and optionally create first token amc org init|score|reportOrg graph and scorecards amc bench init|registry publishBenchmarking registry amc transform init|planTransformation OS amc e2e smokeGo-live smoke tests amc truthguardDeterministic output truth validator amc action-queuePrioritized risk-reduction actions amc confidence-heatmapConfidence heatmap by question amc unknownsKnown unknowns for latest diagnostic amc meta-confidenceConfidence in the score itself
This is a subset. Run amc --help for the complete list, or see the full master reference on GitHub.
Environment Variables
Variable Description Default
AMC_VAULT_PASSPHRASEVault encryption passphrase (non-interactive mode) — AMC_HOMEOverride default .amc workspace location ./.amcAMC_GATEWAY_URLGateway base URL for adapters http://localhost:3210AMC_STUDIO_PORTStudio API/dashboard port 3212AMC_GATEWAY_PORTGateway proxy port 3210AMC_LOG_LEVELLogging verbosity (debug, info, warn, error) infoAMC_TELEMETRYEnable/disable anonymous telemetry offAMC_CICI mode (suppresses interactive prompts) false
Exit Codes
Code Meaning
0Success 1General error 2Invalid arguments or missing required flags 3Workspace not initialized (run amc init) 4Vault locked or passphrase required 5Signature verification failed 10Gate check failed (score below threshold) 20Network/connectivity error
Global Flags
Flag Description
--helpShow help for any command --versionPrint AMC version --jsonOutput as JSON (where supported) --quietSuppress non-essential output --verboseEnable verbose/debug output --dry-runPreview changes without writing (where supported) --agent <id>Target a specific agent --no-colorDisable colored output