feat: add MCP prompts and resources to consent server#81
Open
michaelfarrell76 wants to merge 16 commits intomainfrom
Open
feat: add MCP prompts and resources to consent server#81michaelfarrell76 wants to merge 16 commits intomainfrom
michaelfarrell76 wants to merge 16 commits intomainfrom
Conversation
* feat: add background-task skill and worktree shell helpers Add a background-task skill for parallel worktree workflows, along with safe shell helper functions (worktree-git, worktree-rm, gempty-branches) that are auto-loaded via the existing preToolUse hook. Update the commit-and-push rule to document worktree commands and add .worktrees to .gitignore. * chore: remove gempty-branches helper Not needed for this repo. Remove the shell function and all references from the skill and commit-and-push rule. * style: format SKILL.md
…ivacy-types (#79) * refactor: deduplicate enums and replace inline strings with shared privacy-types - Add CookieOrderField, DataFlowOrderField, DataFlowType, TriageAction, ConsentTrackerType to @transcend-io/privacy-types - Create shared schemas.ts in mcp-server-consent and mcp-server-dsr to deduplicate z.nativeEnum wrappers (ConsentTrackerStatusEnum, OrderDirectionEnum, RequestTypeEnum) - Replace z.string() with z.nativeEnum() for order_field params in consent_list_cookies and consent_list_data_flows - Replace inline z.enum(['APPROVE','JUNK']) and z.enum(['cookie','data_flow']) with shared TriageAction and ConsentTrackerType enums - Replace z.enum(['DRAFT','PUBLISHED']) with AssessmentFormTemplateStatus - Replace z.array(z.string()) scopes with z.array(z.nativeEnum(ScopeName)) - Replace z.string() identifierType with z.nativeEnum(IdentifierType) in preferences tools - Add @transcend-io/privacy-types dependency to mcp-server-admin and mcp-server-preferences * chore: add changeset for privacy-types enum additions * refactor: inline z.nativeEnum, remove shared schemas, enhance admin_create_api_key - Remove schemas.ts from mcp-server-consent and mcp-server-dsr; inline z.nativeEnum() directly in each tool file - Enrich admin_create_api_key tool description with TRANSCEND_SCOPES metadata (title, description, dependency info for every scope) * revert: keep identifierType as z.string() in preferences tools
- Register getConsentPrompts and getConsentResources in cli.ts createMCPServer call (were defined but never wired up) - Export getConsentPrompts and getConsentResources from package index - Remove duplicate .worktrees entry in .gitignore - Remove unnecessary eslint-disable comments - Fix conditional blank lines in triage prompt template
@transcend-io/cli
@transcend-io/privacy-types
@transcend-io/sdk
@transcend-io/utils
@transcend-io/mcp-server
@transcend-io/mcp-server-admin
@transcend-io/mcp-server-assessments
@transcend-io/mcp-server-consent
@transcend-io/mcp-server-core
@transcend-io/mcp-server-discovery
@transcend-io/mcp-server-dsr
@transcend-io/mcp-server-inventory
@transcend-io/mcp-server-preferences
@transcend-io/mcp-server-workflows
commit: |
SKILL.md and docs-reference.md duplicated content that is canonically defined in the MCP prompts and classification-guide resource. Slim the Cursor-specific files to be orchestration wrappers that reference the MCP server as the source of truth for workflow details, research methodology, console commands, and classification guidance. Agents stay self-contained since subagents need complete instructions.
…resources Resources now fetch live content from docs.transcend.io with static fallback: - Tracking Purposes (purpose definitions and taxonomy) - Triage Guide (full classification workflow) - Debugging & Testing (URL overrides, console commands) - Data Flows & Cookies (telemetry and regulation overview) - Telemetry Overview (collection methods) Uses https:// URIs so Cursor can open them directly (fixes "Failed to open resource" error with the custom consent:// scheme).
Skills and agents now delegate to MCP prompts for all substance: - SKILL.md: 117 -> 25 lines. Invokes the consent-triage MCP prompt and documents subagent spawning pattern. - docs-reference.md: deleted. MCP resources + tool discovery replace it. - Inspector agent: 123 -> 35 lines. Fetches consent-inspect-site MCP prompt for methodology. Fixed readonly->false for MCP access. - Researcher agent: 95 -> 40 lines. Fetches consent-research-tracker MCP prompt for methodology. Fixed readonly->false for MCP access.
Member
@michaelfarrell76 indeed there is a better way, and @isaac-martin already set it up! ➡️ https://docs.transcend.io/llms.txt Every doc can be read in markdown by appending https://docs.transcend.io/docs/articles/consent-management/configuration/triage-cookies-and-dataflows-guide |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Note to Reviewer
I could use a deep review on this one - works really well! but still figuring out where to put what.
some questions for reviewers:
Summary
mcp-server-core—createMCPServernow accepts optionalgetPrompts/getResourcesfactories and registersprompts/list,prompts/get,resources/list,resources/readhandlersconsent-triage(full triage workflow),consent-research-tracker(classification research),consent-inspect-site(live site investigation via DevTools)classification-guide(tracking purpose reference)consent-triage) and 2 custom agent definitions for parallel research and site inspectionPromptDefinition,PromptMessage,PromptArgument,ResourceDefinitiontypes exported frommcp-server-coreTest plan
pnpm typecheck— bothmcp-server-coreandmcp-server-consentpassprompts/listreturns 3 promptsresources/listreturns the classification guide resourceprompts/getforconsent-triagewith default args and customtriage_type/batch_sizeresources/readforconsent://classification-guideMade with Cursor