vise --help or vise <command> --help. Each command has an MCP tool equivalent in snake_case.
Most commands take a project path (default: the current directory) and write their state under
sp-vise/. Commands that change generated code are run by your AI agent through the skill; the commands below are the ones you or your reviewer run directly.Explore and plan
| Command | Purpose |
|---|---|
vise explore "<request>" | Map a request to what social.plus offers — no project or API key needed |
vise inspect [path] | Detect platform, monorepo surfaces, design signals, and available sensors |
vise plan [path] --request "..." | Grounded implementation plan with intake questions and docs citations (--summary for a compact view) |
vise init [path] --request "..." --answer key=value | Write the sp-vise/ compliance contract once blocking intake is answered |
| Command | Purpose |
|---|---|
vise workplan next [path] --request "..." | Print the next uncompleted surface and its focused commands |
vise workplan status [path] --request "..." | Show the sequence and which surfaces are done |
vise workplan complete [path] --request "..." --surface <id> | Record a green-checked surface with snapshot evidence |
Validate and verify
| Command | Purpose |
|---|---|
vise check [path] | Validate the code against the recorded contract |
vise check [path] --ci | Read-only; exits non-zero unless green |
vise check [path] --new-only | Gate only on findings introduced since the baseline |
vise check [path] --allow-proof-waiver | Accept an honest runtime-proof-waived result as passing |
vise validate [path] | Run the deterministic validators only (no attestation comparison) |
vise status [path] | Summarize the current compliance state |
vise explain <ruleId> | Print a rule’s rationale, evidence requirements, and remediation |
Evidence
| Command | Purpose |
|---|---|
vise baseline [path] | Snapshot pre-existing findings so --new-only can ignore them |
vise attest [path] --rule <id> --signer host-agent --confidence high --evidence-file evidence.json --rationale "..." | Record that a rule is satisfied through architecture the check can’t see |
vise sync [path] | Persist deterministic-pass evidence after a green check |
vise smoke [path] --log <file> | Assess a captured mount-smoke log into a pass/fail verdict |
vise smoke waive [path] --reason "..." --mode declined|deviceless | Record an auditable runtime-proof waiver |
Docs and SDK facts
| Command | Purpose |
|---|---|
vise search-docs "<query>" | Search social.plus docs for relevant pages |
vise get-doc-page <path> | Fetch a specific doc page by path |
vise sdk-facts --platform <platform> | Read bundled SDK surface facts (--capability, --format json) |
vise debug [path] --error "..." | Diagnose an SDK-specific runtime failure |
Design (advisory)
| Command | Purpose |
|---|---|
vise design extract <prototype> | Extract a graded design contract from an HTML/CSS prototype (--from-project to derive from your own tokens) |
vise design check [path] | Advisory token-conformance report, including computed WCAG for the brand palette |
vise design preview [path] | Write a self-contained visual review for human or VLM judgment |
Skill and MCP
| Command | Purpose |
|---|---|
vise install-skill --target <host> | Install the bundled skill (claude, cursor, vscode, copilot, codex) |
vise print-skill | Print the skill markdown to stdout |
vise mcp | Start the local MCP server |
vise doctor | Verify the install — version, path, docs source |
Related
How Vise works
The loop behind these commands and how to read results.
Run Vise as an MCP server
The MCP tool equivalents of these commands.