Documentation
¶
Index ¶
- Constants
- func CheckAncoraMCPInstalledForTarget(target string) bool
- func CheckMCPInstalled() bool
- func CheckOllama() (installed bool, running bool, path string, err error)
- func CheckVelaMCPInstalledForTarget(target string) bool
- func GetOllamaModels() ([]string, error)
- func InstallOllama() error
- func PullModel(model string) error
- func PullModelSilent(model string) error
- func SaveIntegrationState(state IntegrationState) error
- func StartOllama() error
- func UninstallMCP() error
- type HardwareInfo
- type IntegrationState
- type RecommendedModel
- type RequirementCheck
- type WizardModel
- type WizardStep
Constants ¶
const ( IntegrationModeAncoraOnly = "ancora-only" IntegrationModeAncoraVela = "ancora+vela" IntegrationModeVelaOnly = "vela-only" IntegrationTargetClaudeCode = "claude-code" IntegrationTargetOpenCode = "opencode" IntegrationTargetSkip = "skip" )
const LocalSearchEmbeddingModel = "nomic-embed-text"
LocalSearchEmbeddingModel is the local embedding model Vela expects when it needs to materialize the retrieval substrate for search.
Variables ¶
This section is empty.
Functions ¶
func CheckAncoraMCPInstalledForTarget ¶ added in v1.7.0
func CheckMCPInstalled ¶
func CheckMCPInstalled() bool
CheckMCPInstalled checks if vela is configured in a supported MCP client.
func CheckOllama ¶
CheckOllama checks if Ollama is installed and running.
func CheckVelaMCPInstalledForTarget ¶ added in v1.7.0
func GetOllamaModels ¶
GetOllamaModels returns list of installed Ollama models.
func InstallOllama ¶
func InstallOllama() error
InstallOllama guides installation of Ollama based on platform.
func PullModelSilent ¶
PullModelSilent downloads an Ollama model without output (for TUI mode).
func SaveIntegrationState ¶ added in v1.7.0
func SaveIntegrationState(state IntegrationState) error
func UninstallMCP ¶ added in v1.3.0
func UninstallMCP() error
UninstallMCP removes Vela's MCP registration from supported client configs.
Types ¶
type HardwareInfo ¶
type HardwareInfo struct {
OS string
Arch string
CPUModel string
CPUCores int
TotalRAMGB int
FreeRAMGB int
DiskFreeGB int
}
HardwareInfo represents detected system hardware
func DetectHardware ¶
func DetectHardware() (*HardwareInfo, error)
DetectHardware detects system hardware specs
func (*HardwareInfo) CanRunLocalLLM ¶
func (h *HardwareInfo) CanRunLocalLLM() bool
CanRunLocalLLM checks if system meets minimum requirements for local LLM
func (*HardwareInfo) Format ¶
func (h *HardwareInfo) Format() string
Format returns human-readable hardware summary
func (*HardwareInfo) RecommendedModel ¶
func (h *HardwareInfo) RecommendedModel() string
RecommendedModel returns recommended model based on available RAM
type IntegrationState ¶ added in v1.7.0
type IntegrationState struct {
Mode string `json:"mode"`
MCPTarget string `json:"mcp_target,omitempty"`
UpdatedBy string `json:"updated_by,omitempty"`
UpdatedAt string `json:"updated_at,omitempty"`
}
func LoadIntegrationState ¶ added in v1.7.0
func LoadIntegrationState() (*IntegrationState, error)
type RecommendedModel ¶
type RecommendedModel struct {
Name string
DisplayName string
Description string
Size string
ParameterCount string
QuantizationLevel string // Q4_K_M, Q5_K_M, etc.
// Syfra/Ancora ecosystem compatibility
SyfraCompatible bool // Works with Syfra ecosystem
AncoraEmbedCompat bool // Compatible with Ancora's embedding approach
EmbedDimensions int // Embedding dimensions (768 for nomic-embed-text)
RecommendedFor []string // ["extraction", "reasoning", "embedding"]
// Performance characteristics
MinRAM string // "4GB", "8GB", "16GB"
AvgTokensPerSecond int // Approximate on consumer hardware
ContextWindow int // Token context window
// Model-specific notes
Notes string
SyfraVerified bool // Tested and verified by Syfra team
}
RecommendedModel represents a tested Ollama model with compatibility metadata
func FilterByRAM ¶
func FilterByRAM(models []RecommendedModel, availableGB int) []RecommendedModel
FilterByRAM returns models that fit in available RAM
func FilterBySyfraVerified ¶
func FilterBySyfraVerified(models []RecommendedModel) []RecommendedModel
FilterBySyfraVerified returns only Syfra-tested models
func GetDefaultModel ¶
func GetDefaultModel() RecommendedModel
GetDefaultModel returns the recommended default (llama3)
func GetEmbeddingCompatibleModels ¶
func GetEmbeddingCompatibleModels() []RecommendedModel
GetEmbeddingCompatibleModels returns the embedding model Vela expects for local retrieval/search setup.
func GetRecommendedModels ¶
func GetRecommendedModels() []RecommendedModel
GetRecommendedModels returns curated list of models for Vela/Syfra ecosystem
func (RecommendedModel) FormatModelChoice ¶
func (m RecommendedModel) FormatModelChoice() string
FormatModelChoice formats a model for display in TUI
type RequirementCheck ¶
type RequirementCheck struct {
Name string
Description string
Status string // "checking", "pass", "fail", "warning"
Message string
Required bool
}
RequirementCheck represents a single system requirement validation
type WizardModel ¶
type WizardModel struct {
// contains filtered or unexported fields
}
func NewWizard ¶
func NewWizard() WizardModel
func (WizardModel) FooterHelp ¶
func (m WizardModel) FooterHelp() string
func (WizardModel) Init ¶
func (m WizardModel) Init() tea.Cmd
func (WizardModel) Quitting ¶
func (m WizardModel) Quitting() bool
func (WizardModel) View ¶
func (m WizardModel) View() string
func (WizardModel) ViewContent ¶
func (m WizardModel) ViewContent() string
type WizardStep ¶
type WizardStep int
const ( StepWelcome WizardStep = iota StepSystemCheck StepModeChoice StepProviderChoice StepLocalSetup // Install Ollama + start + pull model StepRemoteSetup // Configure API keys StepMCPConfig StepValidation StepComplete StepError )