samplingtools

package
v1.0.5 Latest Latest
Warning

This package is not in the latest version of its module.

 Go to latest Published: Apr 22, 2026 License: MIT Imports: 25 Imported by: 0

Documentation

Overview

Package samplingtools provides MCP tools that leverage the sampling capability to analyze GitLab data through LLM-assisted summarization and review.

Index

Constants

This section is empty.

Variables

This section is empty.

Functions

func FormatAnalyzeCIConfigMarkdown 

func FormatAnalyzeCIConfigMarkdown(a AnalyzeCIConfigOutput) string

FormatAnalyzeCIConfigMarkdown renders an LLM-generated CI config analysis.

func FormatAnalyzeDeploymentHistoryMarkdown 

func FormatAnalyzeDeploymentHistoryMarkdown(a AnalyzeDeploymentHistoryOutput) string

FormatAnalyzeDeploymentHistoryMarkdown renders an LLM-generated deployment analysis.

func FormatAnalyzeIssueScopeMarkdown 

func FormatAnalyzeIssueScopeMarkdown(a AnalyzeIssueScopeOutput) string

FormatAnalyzeIssueScopeMarkdown renders an LLM-generated issue scope analysis.

func FormatAnalyzeMRChangesMarkdown 

func FormatAnalyzeMRChangesMarkdown(a AnalyzeMRChangesOutput) string

FormatAnalyzeMRChangesMarkdown renders an LLM-generated MR analysis.

func FormatAnalyzePipelineFailureMarkdown 

func FormatAnalyzePipelineFailureMarkdown(a AnalyzePipelineFailureOutput) string

FormatAnalyzePipelineFailureMarkdown renders an LLM-generated pipeline failure analysis.

func FormatCIConfigForAnalysis 

func FormatCIConfigForAnalysis(lint cilint.Output) string

FormatCIConfigForAnalysis builds a Markdown document from CI lint results for LLM configuration analysis.

func FormatDeploymentHistoryForAnalysis 

func FormatDeploymentHistoryForAnalysis(depList deployments.ListOutput, environment string) string

FormatDeploymentHistoryForAnalysis builds a Markdown document from deployment records for LLM deployment pattern analysis.

func FormatFindTechnicalDebtMarkdown 

func FormatFindTechnicalDebtMarkdown(f FindTechnicalDebtOutput) string

FormatFindTechnicalDebtMarkdown renders an LLM-generated technical debt analysis.

func FormatGenerateMilestoneReportMarkdown 

func FormatGenerateMilestoneReportMarkdown(r GenerateMilestoneReportOutput) string

FormatGenerateMilestoneReportMarkdown renders an LLM-generated milestone report.

func FormatGenerateReleaseNotesMarkdown 

func FormatGenerateReleaseNotesMarkdown(r GenerateReleaseNotesOutput) string

FormatGenerateReleaseNotesMarkdown renders LLM-generated release notes.

func FormatIssueForSummary 

func FormatIssueForSummary(issue issues.Output, notes issuenotes.ListOutput) string

FormatIssueForSummary builds a Markdown document from an issue and its notes, suitable for passing to an LLM for discussion summarization.

func FormatIssueScopeForAnalysis 

func FormatIssueScopeForAnalysis(issue issues.Output, timeStats issues.TimeStatsOutput, participants issues.ParticipantsOutput, closingMRs, relatedMRs issues.RelatedMRsOutput, notes issuenotes.ListOutput) string

FormatIssueScopeForAnalysis builds a Markdown document from issue details, time stats, participants, related MRs, and notes for scope analysis.

func FormatMRForAnalysis 

func FormatMRForAnalysis(mr mergerequests.Output, changes mrchanges.Output) string

FormatMRForAnalysis builds a Markdown document from a merge request and its file diffs, suitable for passing to an LLM for code review analysis. Also used by ReviewMRSecurity for security-focused reviews.

func FormatMRReviewForAnalysis 

func FormatMRReviewForAnalysis(mr mergerequests.Output, discussions mrdiscussions.ListOutput, approvals mrapprovals.StateOutput) string

FormatMRReviewForAnalysis builds a Markdown document from an MR, its discussions, and approval state for LLM review summarization.

func FormatMilestoneForAnalysis 

func FormatMilestoneForAnalysis(ms milestones.Output, msIssues milestones.MilestoneIssuesOutput, msMRs milestones.MilestoneMergeRequestsOutput) string

FormatMilestoneForAnalysis builds a Markdown document from a milestone, its issues, and merge requests for LLM report generation.

func FormatPipelineFailureForAnalysis 

func FormatPipelineFailureForAnalysis(pipeline pipelines.DetailOutput, traces []JobTrace) string

FormatPipelineFailureForAnalysis builds a Markdown document from pipeline details and failed job traces for LLM failure analysis.

func FormatReleaseDataForAnalysis 

func FormatReleaseDataForAnalysis(from, to string, cmp repository.CompareOutput, mrs mergerequests.ListOutput) string

FormatReleaseDataForAnalysis builds a Markdown document combining commits, merged MRs, and file diffs, suitable for LLM release notes generation.

func FormatReviewMRSecurityMarkdown 

func FormatReviewMRSecurityMarkdown(r ReviewMRSecurityOutput) string

FormatReviewMRSecurityMarkdown renders an LLM-generated MR security review.

func FormatSummarizeIssueMarkdown 

func FormatSummarizeIssueMarkdown(s SummarizeIssueOutput) string

FormatSummarizeIssueMarkdown renders an LLM-generated issue summary.

func FormatSummarizeMRReviewMarkdown 

func FormatSummarizeMRReviewMarkdown(s SummarizeMRReviewOutput) string

FormatSummarizeMRReviewMarkdown renders an LLM-generated MR review summary.

func FormatTechnicalDebtForAnalysis 

func FormatTechnicalDebtForAnalysis(blobs []search.BlobOutput) string

FormatTechnicalDebtForAnalysis builds a Markdown document from code search results containing debt markers for LLM analysis.

func RegisterMeta added in v1.0.4 

func RegisterMeta(server *mcp.Server, client *gitlabclient.Client)

RegisterMeta registers a single gitlab_analyze meta-tool that consolidates all 11 sampling analysis tools under one action-dispatched interface.

func RegisterTools 

func RegisterTools(server *mcp.Server, client *gitlabclient.Client)

RegisterTools wires sampling-powered tools to the MCP server.

func SamplingUnsupportedResult 

func SamplingUnsupportedResult(toolName string) *mcp.CallToolResult

SamplingUnsupportedResult returns a structured error tool result when the MCP client does not support sampling. Suggests alternative non-sampling tools.

Types

type AnalyzeCIConfigInput 

type AnalyzeCIConfigInput struct {
	ProjectID  toolutil.StringOrInt `json:"project_id"  jsonschema:"Project ID or URL-encoded path,required"`
	ContentRef string               `json:"content_ref" jsonschema:"Branch or tag for CI content (defaults to default branch)"`
}

AnalyzeCIConfigInput defines parameters for LLM-assisted CI configuration analysis.

type AnalyzeCIConfigOutput 

type AnalyzeCIConfigOutput struct {
	toolutil.HintableOutput
	ProjectID string `json:"project_id"`
	Valid     bool   `json:"valid"`
	Analysis  string `json:"analysis"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

AnalyzeCIConfigOutput holds the LLM analysis of a CI configuration.

func AnalyzeCIConfig 

func AnalyzeCIConfig(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input AnalyzeCIConfigInput) (AnalyzeCIConfigOutput, error)

AnalyzeCIConfig lints a project's CI configuration and delegates to the MCP sampling capability for LLM-assisted analysis of the configuration quality.

type AnalyzeDeploymentHistoryInput 

type AnalyzeDeploymentHistoryInput struct {
	ProjectID   toolutil.StringOrInt `json:"project_id"            jsonschema:"Project ID or URL-encoded path,required"`
	Environment string               `json:"environment,omitempty" jsonschema:"Filter by environment name (e.g. production, staging)"`
}

AnalyzeDeploymentHistoryInput defines parameters for LLM-assisted deployment analysis.

type AnalyzeDeploymentHistoryOutput 

type AnalyzeDeploymentHistoryOutput struct {
	toolutil.HintableOutput
	ProjectID   string `json:"project_id"`
	Environment string `json:"environment,omitempty"`
	Analysis    string `json:"analysis"`
	Model       string `json:"model"`
	Truncated   bool   `json:"truncated"`
}

AnalyzeDeploymentHistoryOutput holds the LLM analysis of deployment history.

func AnalyzeDeploymentHistory 

func AnalyzeDeploymentHistory(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input AnalyzeDeploymentHistoryInput) (AnalyzeDeploymentHistoryOutput, error)

AnalyzeDeploymentHistory fetches recent deployments, then delegates to the MCP sampling capability for deployment pattern analysis.

type AnalyzeIssueScopeInput 

type AnalyzeIssueScopeInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path,required"`
	IssueIID  int64                `json:"issue_iid"  jsonschema:"Issue internal ID,required"`
}

AnalyzeIssueScopeInput defines parameters for LLM-assisted issue scope analysis.

type AnalyzeIssueScopeOutput 

type AnalyzeIssueScopeOutput struct {
	toolutil.HintableOutput
	IssueIID  int64  `json:"issue_iid"`
	Title     string `json:"title"`
	Analysis  string `json:"analysis"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

AnalyzeIssueScopeOutput holds the LLM analysis of an issue's scope and effort.

func AnalyzeIssueScope 

func AnalyzeIssueScope(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input AnalyzeIssueScopeInput) (AnalyzeIssueScopeOutput, error)

AnalyzeIssueScope fetches an issue with its time stats, participants, related MRs, and notes, then delegates to the MCP sampling capability for scope analysis.

type AnalyzeMRChangesInput 

type AnalyzeMRChangesInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path"`
	MRIID     int64                `json:"mr_iid"     jsonschema:"Merge request internal ID"`
}

AnalyzeMRChangesInput defines parameters for LLM-assisted MR code review.

type AnalyzeMRChangesOutput 

type AnalyzeMRChangesOutput struct {
	toolutil.HintableOutput
	MRIID     int64  `json:"mr_iid"`
	Title     string `json:"title"`
	Analysis  string `json:"analysis"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

AnalyzeMRChangesOutput holds the LLM analysis result for an MR.

func AnalyzeMRChanges 

func AnalyzeMRChanges(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input AnalyzeMRChangesInput) (AnalyzeMRChangesOutput, error)

AnalyzeMRChanges fetches a merge request and its diffs, then delegates to the MCP client's sampling capability for LLM-assisted code review analysis. Returns sampling.ErrSamplingNotSupported if the client lacks sampling support.

type AnalyzePipelineFailureInput 

type AnalyzePipelineFailureInput struct {
	ProjectID  toolutil.StringOrInt `json:"project_id"  jsonschema:"Project ID or URL-encoded path,required"`
	PipelineID int64                `json:"pipeline_id" jsonschema:"Pipeline ID to analyze,required"`
}

AnalyzePipelineFailureInput defines parameters for the analyze pipeline failure operation.

type AnalyzePipelineFailureOutput 

type AnalyzePipelineFailureOutput struct {
	toolutil.HintableOutput
	PipelineID int64  `json:"pipeline_id"`
	Status     string `json:"status"`
	Ref        string `json:"ref"`
	Analysis   string `json:"analysis"`
	Model      string `json:"model"`
	Truncated  bool   `json:"truncated"`
}

AnalyzePipelineFailureOutput holds the LLM analysis of a failed pipeline.

func AnalyzePipelineFailure 

func AnalyzePipelineFailure(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input AnalyzePipelineFailureInput) (AnalyzePipelineFailureOutput, error)

AnalyzePipelineFailure fetches pipeline details, failed jobs and their traces, then delegates to the MCP sampling capability for failure root cause analysis.

type FindTechnicalDebtInput 

type FindTechnicalDebtInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path,required"`
	Ref       string               `json:"ref"        jsonschema:"Branch or tag to search in (defaults to default branch)"`
}

FindTechnicalDebtInput defines parameters for LLM-assisted technical debt discovery.

type FindTechnicalDebtOutput 

type FindTechnicalDebtOutput struct {
	toolutil.HintableOutput
	ProjectID string `json:"project_id"`
	Analysis  string `json:"analysis"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

FindTechnicalDebtOutput holds the LLM analysis of technical debt indicators.

func FindTechnicalDebt 

func FindTechnicalDebt(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input FindTechnicalDebtInput) (FindTechnicalDebtOutput, error)

FindTechnicalDebt searches for debt markers (TODO, FIX-ME, HACK, XXX, DEPRECATED) in the project codebase, then delegates to the MCP sampling capability for analysis.

type GenerateMilestoneReportInput 

type GenerateMilestoneReportInput struct {
	ProjectID    toolutil.StringOrInt `json:"project_id"    jsonschema:"Project ID or URL-encoded path,required"`
	MilestoneIID int64                `json:"milestone_iid" jsonschema:"Milestone IID (project-scoped),required"`
}

GenerateMilestoneReportInput defines parameters for LLM-assisted milestone reporting.

type GenerateMilestoneReportOutput 

type GenerateMilestoneReportOutput struct {
	toolutil.HintableOutput
	MilestoneIID int64  `json:"milestone_iid"`
	Title        string `json:"title"`
	Report       string `json:"report"`
	Model        string `json:"model"`
	Truncated    bool   `json:"truncated"`
}

GenerateMilestoneReportOutput holds the LLM-generated milestone report.

func GenerateMilestoneReport 

func GenerateMilestoneReport(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input GenerateMilestoneReportInput) (GenerateMilestoneReportOutput, error)

GenerateMilestoneReport fetches milestone details, its issues and MRs, then delegates to the MCP sampling capability for report generation.

type GenerateReleaseNotesInput 

type GenerateReleaseNotesInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path,required"`
	From      string               `json:"from"       jsonschema:"Starting ref: tag, branch or SHA,required"`
	To        string               `json:"to"         jsonschema:"Ending ref: tag, branch or SHA (defaults to HEAD)"`
}

GenerateReleaseNotesInput defines parameters for LLM-assisted release notes generation.

type GenerateReleaseNotesOutput 

type GenerateReleaseNotesOutput struct {
	toolutil.HintableOutput
	From         string `json:"from"`
	To           string `json:"to"`
	ReleaseNotes string `json:"release_notes"`
	Model        string `json:"model"`
	Truncated    bool   `json:"truncated"`
}

GenerateReleaseNotesOutput holds the LLM-generated release notes.

func GenerateReleaseNotes 

func GenerateReleaseNotes(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input GenerateReleaseNotesInput) (GenerateReleaseNotesOutput, error)

GenerateReleaseNotes fetches commits, diffs, and merged MRs between two refs, then delegates to the MCP sampling capability for LLM-generated release notes. Returns sampling.ErrSamplingNotSupported if the client lacks sampling support.

type IssueContextResult 

type IssueContextResult struct {
	IID     int64
	Title   string
	Content string // Formatted Markdown ready for LLM
}

IssueContextResult holds the GraphQL-fetched issue context.

func BuildIssueContext 

func BuildIssueContext(ctx context.Context, client *gitlabclient.Client, projectPath string, issueIID int64) (IssueContextResult, error)

BuildIssueContext fetches issue details, notes, participants, time tracking, labels, assignees, milestone, and related MRs in a single GraphQL request. This replaces up to 6 sequential REST calls used by analyze_issue_scope.

type JobTrace 

type JobTrace struct {
	Job   jobs.Output
	Trace string
}

JobTrace pairs a job with its trace output for pipeline failure analysis.

type MRContextResult 

type MRContextResult struct {
	IID     int64
	Title   string
	Content string // Formatted Markdown ready for LLM
}

MRContextResult holds the GraphQL-fetched merge request context.

func BuildMRContext 

func BuildMRContext(ctx context.Context, client *gitlabclient.Client, projectPath string, mrIID int64) (MRContextResult, error)

BuildMRContext fetches MR details, discussions, approval state, and head pipeline status in a single GraphQL request and formats it as Markdown. The projectPath must be the full path (e.g. "group/project"), not a numeric ID.

type PipelineContextResult 

type PipelineContextResult struct {
	PipelineIID  int64
	Status       string
	Ref          string
	Content      string  // Formatted Markdown ready for LLM
	FailedJobIDs []int64 // Job IDs of failed jobs (for REST trace fetching)
}

PipelineContextResult holds the GraphQL-fetched pipeline context.

func BuildPipelineContext 

func BuildPipelineContext(ctx context.Context, client *gitlabclient.Client, projectPath string, pipelineIID int64) (PipelineContextResult, error)

BuildPipelineContext fetches pipeline details with stages and jobs in a single GraphQL request. Job traces are not available via GraphQL and must be fetched separately via REST. The returned FailedJobIDs can be used for trace retrieval.

type ReviewMRSecurityInput 

type ReviewMRSecurityInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path,required"`
	MRIID     int64                `json:"mr_iid"     jsonschema:"Merge request internal ID,required"`
}

ReviewMRSecurityInput defines parameters for LLM-assisted MR security review.

type ReviewMRSecurityOutput 

type ReviewMRSecurityOutput struct {
	toolutil.HintableOutput
	MRIID     int64  `json:"mr_iid"`
	Title     string `json:"title"`
	Review    string `json:"review"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

ReviewMRSecurityOutput holds the LLM security review of an MR.

func ReviewMRSecurity 

func ReviewMRSecurity(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input ReviewMRSecurityInput) (ReviewMRSecurityOutput, error)

ReviewMRSecurity fetches an MR and its diffs, then delegates to the MCP sampling capability for a security-focused code review.

type SummarizeIssueInput 

type SummarizeIssueInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path"`
	IssueIID  int64                `json:"issue_iid"  jsonschema:"Issue internal ID"`
}

SummarizeIssueInput defines parameters for LLM-assisted issue summarization.

type SummarizeIssueOutput 

type SummarizeIssueOutput struct {
	toolutil.HintableOutput
	IssueIID  int64  `json:"issue_iid"`
	Title     string `json:"title"`
	Summary   string `json:"summary"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

SummarizeIssueOutput holds the LLM summary of an issue.

func SummarizeIssue 

func SummarizeIssue(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input SummarizeIssueInput) (SummarizeIssueOutput, error)

SummarizeIssue fetches an issue and its notes, then delegates to the MCP client's sampling capability for LLM-assisted summarization of the discussion. Returns sampling.ErrSamplingNotSupported if the client lacks sampling support.

type SummarizeMRReviewInput 

type SummarizeMRReviewInput struct {
	ProjectID toolutil.StringOrInt `json:"project_id" jsonschema:"Project ID or URL-encoded path,required"`
	MRIID     int64                `json:"mr_iid"     jsonschema:"Merge request internal ID,required"`
}

SummarizeMRReviewInput defines parameters for LLM-assisted MR review summarization.

type SummarizeMRReviewOutput 

type SummarizeMRReviewOutput struct {
	toolutil.HintableOutput
	MRIID     int64  `json:"mr_iid"`
	Title     string `json:"title"`
	Summary   string `json:"summary"`
	Model     string `json:"model"`
	Truncated bool   `json:"truncated"`
}

SummarizeMRReviewOutput holds the LLM summary of an MR review.

func SummarizeMRReview 

func SummarizeMRReview(ctx context.Context, req *mcp.CallToolRequest, client *gitlabclient.Client, input SummarizeMRReviewInput) (SummarizeMRReviewOutput, error)

SummarizeMRReview fetches an MR, its discussions, and approval state, then delegates to the MCP sampling capability for review summarization.

Source Files

View all Source files

Jump to

Keyboard shortcuts

? : This menu
/ : Search site
f or F : Jump to
y or Y : Canonical URL
go.dev uses cookies from Google to deliver and enhance the quality of its services and to analyze traffic. Learn more.