MCP Tools Reference

ClarityLoop MCP exposes tools for reading and updating people-development data. Use this page to understand what an AI client or agent can access when you build dashboards, reports, workflow assistants, review prep, manager copilots, or custom analytics on top of the ClarityLoop MCP server.

MCP clients call tools, not REST endpoints. Each tool returns JSON and runs as the connected ClarityLoop user.

Capabilities at a Glance

Area	Read tools	Write tools	Useful for
People and org context	getCurrentUser, getPeopleFilterOptions, searchPeople, getDirectReports	-	identity lookup, team scoping, segmentation, manager dashboards
Feedback and growth signals	findReceivedFeedback, findGivenFeedback, findIncomingFeedbackRequests, findSentFeedbackRequests, getFeedbackDetails, getGrowthSignals	giveFeedback, requestFeedback	feedback timelines, review prep, growth summaries, feedback operations
Work context and personal notes	getWorkSummary, searchNotes	saveNote	evidence gathering, self-review prep, manager notes, reflection workflows
Goals and OKRs	findGoals, getGoalDetails	createGoal, linkToGoal, addGoalItem, updateGoalProgress	OKR dashboards, growth plans, goal hygiene, progress nudges
Meetings and 1:1s	findMeetings, getMeetingDetails	scheduleMeeting, addMeetingAgendaItem, addMeetingActionItem, addMeetingNote	1:1 prep, meeting summaries, action tracking, agenda automation
Career framework and values	getCareerFramework, getLevelDetails, listCompanyValues	-	competency maps, leveling support, value-based feedback
Leadership insights	getLeadershipInsights, findPeopleByInsightCriteria	-	team health dashboards, manager focus areas, risk signals
Surveys	findSurveys, getSurveyInsights	-	engagement dashboards, survey trend reports, theme analysis
Knowledge	searchKnowledge	-	product help, people-science guidance, agent grounding

How Scope Works

MCP access follows the connected user's ClarityLoop permissions. The MCP API key does not create a new workspace-wide permission layer; it lets an AI client act with the same access as the person who connected it.

Workspace Members

Workspace members can use MCP for their own ClarityLoop context.

That can include:

• their profile, role, level, affiliations, and member attributes
• their own goals and growth signals
• feedback they have received or given, subject to feedback visibility settings
• feedback requests assigned to them
• their meetings, agendas, action items, and meeting notes
• their personal notes
• work context from connected systems and auto-capture where available

Managers

Managers can use MCP across their reporting line.

That can include:

• direct-report lists and reporting-line context
• line-report profiles, roles, levels, affiliations, and member attributes
• line-report goals, growth signals, and received feedback where manager visibility allows it
• 1:1s, agendas, notes, and action items involving their line reports
• leadership insight sections scoped to their reporting line
• insight criteria queries for people in their reporting line
• survey insights where they are a survey leader and the results are scoped to their line

Managers cannot use owner-only workspace filters unless they are also workspace owners.

Workspace Owners

Workspace owners can use broader workspace-level access where MCP tools support it.

That can include:

• people search and segmentation across the workspace
• leadership insights filtered by leader, department, role, affiliation, or member attribute
• insight criteria queries across selected workspace segments
• survey insights and survey filters for surveys they own
• workspace career framework and company values

Owner access is still governed by ClarityLoop product rules.

Write Access

Write tools create or update real ClarityLoop records. Agent workflows should show the intended change and ask for confirmation before calling tools that create feedback, request feedback, schedule meetings, add agenda items, add action items, add notes, create goals, link goals, or update progress.

People and Org Context

getCurrentUser

Returns the connected user's ClarityLoop profile and workspace context.

Returned data includes:

• user id, email, name
• workspace id, name
• job title
• workspaceMemberRole
• current level, role, track, and department
• affiliations
• member attribute values, including whether an attribute is marked sensitive

Use this as the first tool in most agent workflows to understand the acting user, workspace, and permission context.

getPeopleFilterOptions

Returns the segmentation options available in the workspace. Use this when an agent or application needs to ask questions like:

• "show this for Engineering"
• "compare managers only"
• "filter survey results by department"
• "look at people in this role, level, location, cohort, or custom attribute"

The tool tells the agent which people filters exist in this workspace. That can include departments, roles, levels, affiliations, and configured member attributes such as location, employment type, cohort, or other workspace-specific fields.

Returned data includes:

• career framework context: the published framework and the department tree used for role and level filters
• departments: department names and their available roles, tracks, and levels
• Affiliations: workspace-defined teams, projects, or similar groupings
• member attributes: configurable people attributes and their available values, including whether an attribute is marked sensitive
• search filters: additional people-search filters exposed by ClarityLoop, with display names and selectable options

Each option includes the stable identifier the MCP client needs for follow-up tool calls. In a dashboard or app, show the human-readable names; keep the IDs or keys in the background when calling filtered tools.

Use these options to narrow people search, leadership insights, insight criteria, and survey insights. Leader filters are different: use searchPeople to find the manager or leader first, then use that person as the leader filter.

searchPeople

Finds workspace members by name or filters.

Inputs include:

• searchTerm
• includeInactive
• managerOnly
• filters for department, role, level, affiliation, and member attributes
• pagination

pageSize is capped at 20.

Returned data includes:

• user id, email, name
• title and workspace role
• level, role, track, and department
• affiliations
• member attribute values
• active and current

Field notes:

• workspaceMemberRole identifies whether the person is a workspace member or workspace owner.
• active tells you whether the workspace member is currently active.
• current marks the connected user when they appear in search results.
• member attribute values can include sensitive attributes; check the sensitivity flag before displaying or storing them downstream.

Use this before calling person-specific tools such as feedback, goals, growth signals, work context, or direct reports.

Use getPeopleFilterOptions first when the user wants to filter by department, role, level, affiliation, or another workspace-specific people attribute.

getDirectReports

Lists direct reports for a manager. If no userId is provided, the current user is used.

Returned data uses the same person shape as searchPeople.

Use this to create manager views, reporting-line filters, 1:1 prep lists, and team-level summaries.

Feedback and Growth Signals

findReceivedFeedback

Finds feedback received by a specific person.

Inputs include:

• userId
• optional sender fromUserId
• fromDate, toDate
• pagination

pageSize is capped at 20.

Returned data includes:

• feedback id, summary, status
• sender details where visible
• anonymous
• sentimentScore
• feedback request ID
• sharing flags for manager visibility
• linked values, strengths, and growth opportunities
• createdAt, updatedAt

Field notes:

• feedbackRequestId links the feedback back to the request that prompted it. If it is empty, the feedback was given directly rather than as a response to a request.
• anonymous means the giver's identity may be hidden depending on the viewer's permissions.
• sharedWithManager means the recipient's manager can see the feedback.
• sharedOnlyWithManager means the feedback was shared with the manager instead of the recipient.
• sentimentScore is a directional signal for the tone of the feedback; use it for trends and summaries, not as a standalone performance judgment.
• linked strengths, growth opportunities, and values are the ClarityLoop growth and culture signals connected to the feedback.

Useful for feedback timelines, review evidence, sentiment tracking, and growth pattern analysis.

findGivenFeedback

Finds feedback given by the current user.

Inputs include:

• optional recipient toUserId
• fromDate, toDate
• pagination

pageSize is capped at 20.

Returned data includes the same summary shape as received feedback, with recipient details.

Useful for feedback habits, pending review evidence, and manager coaching workflows.

getFeedbackDetails

Returns the full record for one feedback item.

Returned data includes:

• summary and full description
• sender and recipient details where visible
• anonymous and sharing settings
• sentiment score
• linked values, strengths, and growth opportunities
• structured feedback formats: SBI, SSC, STAR
• evidence contexts with links and metadata
• status and timestamps

Field notes:

• SBI, SSC, and STAR are structured feedback formats used to break feedback into reusable evidence.
• evidence contexts are links or records that ground the feedback in actual work.

Use this when a dashboard or agent needs source detail behind a feedback summary.

findIncomingFeedbackRequests

Finds feedback requests assigned to the current user.

Inputs include:

• optional forUserId
• fromDate, toDate
• pagination

pageSize is capped at 20.

Returned data includes request id, type, share ID, summary, status, response count, who the request is about, recipients, response flags, manager flags, and timestamps.

Field notes:

• shareId identifies the shareable request flow when one exists.
• responseCount shows how many people have responded.
• recipient responded flags let you build follow-up and reminder workflows.
• fromManager marks requests sent in the manager context.

Useful for feedback request inboxes and nudges.

findSentFeedbackRequests

Finds feedback requests sent by the current user about a specific person.

Inputs include:

• forUserId
• optional recipient recipientUserId
• optional fromManager
• fromDate, toDate
• pagination

pageSize is capped at 20.

Returned data uses the same request shape as incoming requests.

Field notes:

• Use request status, responseCount, and recipient response flags to show collection progress.
• fromManager separates ordinary feedback requests from requests made by a manager about their line report.

Useful for tracking feedback collection progress during reviews or growth cycles.

getGrowthSignals

Returns a consolidated growth-signal summary for a person.

Inputs include:

• optional userId; defaults to the current user
• optional fromDate, toDate

Returned data includes:

• strengths
• growth areas
• linked company values
• point-in-time sentiment score
• overall sentiment score
• feedback counts given and received

Useful for growth dashboards, review prep, coaching summaries, and focus-area recommendations.

Write Tools

giveFeedback creates or updates a feedback draft and can optionally share it. It supports summary, description, SBI, SSC, STAR, anonymity, manager-sharing flags, feedback-request linkage, value tags, and evidence links.

requestFeedback sends feedback requests to one or more people about a person. It supports request summary, description, manager mode, and evidence links.

Work Context and Personal Notes

getWorkSummary

Returns work context that ClarityLoop has captured from the tools your workspace uses.

This can include supported context from Jira, Confluence, Asana, Linear, Forecast, GitHub, GitLab, Bitbucket, Azure DevOps, HubSpot, and browser-extension auto-capture.

Use it to help an agent ground feedback, goals, growth signals, review prep, and 1:1s in work that happened. Manually captured evidence and browser-extension captures are searched through searchNotes.

Inputs include:

• optional userId; defaults to the current user
• optional collaborator user IDs
• optional search term
• pagination

pageSize is capped at 20.

Returned data includes context item id, link, metadata, createdAt, and updatedAt.

Field notes:

• collaborator filters narrow the feed to shared work involving selected people.
• link points back to the source context where available.
• metadata contains source-specific context about the work item, such as title, source system, or captured details.
• Access is scoped in ClarityLoop. Work context is not exposed as an open workspace activity feed.

Useful for self-assessments, feedback evidence, 1:1 prep, review packets, and manager summaries.

searchNotes

Searches the connected user's personal ClarityLoop notes.

Use this for manually captured evidence, reflections, prep notes, and browser-extension captures that were saved as notes.

Inputs include:

• searchTerm
• collectionId
• pagination

pageSize is capped at 20.

Returned data includes note id, title, description, collection ID, contexts, and timestamps.

These are personal notes, not meeting notes.

saveNote

Creates or updates a personal note.

Inputs include:

• optional noteId
• title
• content

Returned data includes the saved note record.

Useful for capturing prep notes, evidence summaries, coaching reflections, and follow-up records.

Goals and OKRs

findGoals

Lists goals for one person.

Inputs include:

• userId
• optional keyword search
• date range
• statuses
• linked record filter
• pagination

pageSize is capped at 20.

Goal types include deliverables, growth areas, and deliverables plus growth.

Returned data includes:

• goal id, title, status, type, dates, current score
• owner
• parent/cascade data
• links to values, feedback, growth opportunities, or career levels
• key results or actions with IDs, titles, measurement type, current value, status, dates, and source link

Field notes:

• currentScore is the goal's current progress score.
• objectiveType tells you whether the goal is about deliverables, growth, or both.
• links connect the goal to supporting records such as feedback, company values, growth opportunities, or career framework levels.
• key results and actions are returned together because ClarityLoop uses the goal type to decide how each item is interpreted.

Useful for OKR dashboards, growth-plan views, review prep, and goal-health checks.

getGoalDetails

Returns full details for a goal or key result/action.

For goals, returned data includes description, score, links, key results/actions, and comments.

For key results/actions, returned data includes description, measurement type, current value, status, dates, source link, source metadata, owner, and comments.

Write Tools

createGoal creates a goal for a person with title, type, date range, description, and draft or published status.

addGoalItem adds a key result or action to an existing goal.

linkToGoal links a goal to a company value, feedback item, growth opportunity, or career level.

updateGoalProgress updates the progress value for a published key result or action and can add a comment.

Meetings and 1:1s

findMeetings

Lists meetings visible to the connected user in a date range. For a manager, this is useful for finding their own upcoming or past 1:1s with line reports. For an individual, it can support their own meeting prep and follow-up.

It does not let a user browse another person's calendar. To find a meeting with someone else, filter the connected user's meetings by that person as an attendee.

Inputs include:

• startDate, endDate
• includeCancelled
• optional other attendee, to narrow the list to meetings with a specific person

The date range can be up to 45 days.

Returned data includes:

• meeting reference
• meeting ID and parent event ID
• title, event type, location
• start and end time
• time zone
• cancellation state
• recurrence rule
• agenda items
• attendees and declined flags

Field notes:

• meeting reference is the value an agent uses when it needs to add an agenda item, action item, or note to the right meeting.
• meeting ID identifies a specific meeting occurrence when one exists.
• parent event ID identifies the recurring meeting series.
• eventType separates general meetings from 1:1s.
• agenda item crossed indicates whether an agenda item has been marked complete.

getMeetingDetails

Returns available details for one meeting.

Returned data includes the fields from findMeetings, plus meeting notes and action items where available.

Action items include status, due date, assignee, additional assignees, and possible assignees.

Write Tools

scheduleMeeting creates a new meeting. It supports attendee user IDs, title, start and end time, location, time zone, recurrence rule, and meeting type. One-to-one meetings must be between a manager and their direct report.

addMeetingAgendaItem adds an agenda item with an optional evidence link.

addMeetingActionItem adds an action item with assignee, additional assignees, due date, and status.

addMeetingNote adds a meeting note and supports private notes.

Career Framework and Values

getCareerFramework

Returns the published career framework summary and department structure.

Returned data includes:

• framework id, name, workspace ID, published state
• departments with roles, tracks, and levels
• overview text
• private flags

Field notes:

• private flags indicate parts of the framework that are marked private in ClarityLoop.
• departments, roles, tracks, and levels form the structure used when filtering people by career framework.

Useful for competency maps, leveling navigation, growth planning, and role-based analytics.

getLevelDetails

Returns detailed expectations for one level.

Returned data includes:

• level name, sequence, overview, private flag
• responsibilities
• competencies, skills, and behaviours
• related role and track

Useful for role-play, growth opportunity mapping, promotion readiness, and competency-based development plans.

listCompanyValues

Returns published company values with IDs, names, and descriptions.

Useful for tagging feedback, value distribution reporting, and culture-focused coaching.

Leadership Insights

Leadership Insights tools return people analytics for a defined group. For managers, that usually means their reporting line. For workspace owners, it can mean a selected leader tree or workspace segment.

There are two MCP tools in this area:

• getLeadershipInsights returns metric sections for a scope.
• findPeopleByInsightCriteria runs a rule-style insight query and returns the people who match.

getLeadershipInsights

Returns leadership insight sections for the group being analyzed, such as a manager's reporting line or a workspace-owner-selected leader tree or segment.

Available sections:

• sentiment_scores: sentiment score by person
• goal_progress: goal progress score by person
• platform_activity: activity counts across ClarityLoop areas
• focus_areas: recurring growth or coaching themes with signal counts
• value_distribution: company values appearing in feedback and related signals

Inputs include:

• optional sections; if omitted, all sections are returned
• custom startDate and endDate; if omitted, MCP uses the last six months
• owner-only filters: leader, department, role, affiliation, member attribute

If using a custom date range, provide both startDate and endDate.

Use getPeopleFilterOptions for workspace filters and searchPeople for leader filters.

Returned data includes:

• requested section names
• sentiment scores with person ID, name, and score
• goal progress scores with person ID, name, and score
• platform activity metrics with activity type and count
• focus areas with ID, title, description, active state, creation date, feedback count, comment count, and event count
• value distribution rows with value and person context

Field notes:

• sentiment and goal progress scores are useful for trend and cohort analysis; avoid treating one score as a complete assessment of a person.
• platform activity is volume by type. It can include activity such as feedback, feedback requests, meetings, meeting agendas, meeting notes, meeting actions, completed actions, goals updated, notes, AI chat conversations, growth opportunities, strengths, focus areas, work context, and survey responses.
• focus areas are recurring growth or leadership themes detected across ClarityLoop signals. Counts show signal volume, not severity by themselves.
• value distribution depends on published company values and value-tagged feedback.

Access notes:

• Workspace owners can filter by leader, department, role, affiliation, or member attribute.
• Managers are scoped to their own reporting line.
• Non-owners without direct reports receive empty insight sections.

Useful for team health dashboards, manager focus areas, leadership packs, and people-risk follow-up.

findPeopleByInsightCriteria

Finds people who match a set of Leadership Insights criteria. This is useful for report-style questions such as "who has low sentiment and low goal progress?" or "who is giving a lot of developmental feedback?"

Inputs include:

• criteria with signal, comparison, and numeric value
• custom startDate and endDate; if omitted, MCP uses the last six months
• owner-only filters: leader, department, role, affiliation, member attribute

If using a custom date range, provide both startDate and endDate.

Use getPeopleFilterOptions for workspace filters and searchPeople for leader filters.

Supported signals include:

• sentiment score: how a person is being experienced across analyzed signals
• goal progress: progress across goals in the selected period
• feedback received: count of feedback received
• feedback given: count of feedback given
• feedback given with growth opportunities: count of developmental feedback given
• 1:1s: count of structured 1:1 meetings with notes
• completed actions percentage: follow-through on meeting action items
• current level: assigned career framework level sequence
• promotions: promotion count in the selected period

Supported comparisons are EQ, GT, GE, LT, and LE.

Multiple criteria are combined with AND logic. A person must match every criterion to be returned.

Returned data includes matched people with identity, title, level, affiliations, member attributes, active/current flags, workspace role, and creation date.

Field notes:

• Sentiment and goal progress values are numeric scores, typically used on a 0-10 scale.
• Count-based signals such as feedback, 1:1s, and promotions use counts in the selected period.
• current level uses the numeric level sequence from the career framework, not the person's job title.
• completed actions percentage is a percentage-style follow-through metric for meeting action items.

Useful for alerting, cohorts, manager prompts, and targeted follow-up lists.

Surveys

findSurveys

Finds surveys the connected user owns or leads, along with each time the survey has been sent or scheduled. This helps an agent choose the right survey period before fetching results.

Inputs include:

• optional title keyword

Returned data includes:

• survey ID and title
• trigger type
• status
• start date
• survey sends or cycles, sorted newest first
• for each send or cycle: start time, recipient count, responded count, and response rate

Field notes:

• triggerType tells you whether the survey is manually launched or tied to a lifecycle event such as onboarding or offboarding.
• status tells you whether the survey is active, draft, retired, cancelled, or winding down.

Use this first when a workflow needs to pick a survey and choose which survey period to analyze.

getSurveyInsights

Returns aggregated results for a survey period. Use it to understand participation, question-level results, sentiment, and recurring themes when analyzing survey outcomes or planning follow-up.

Inputs include:

• surveyId
• optional surveyRunIds; defaults to the latest 10 runs
• owner-only filters: leader, department, role, affiliation, member attribute

Use getPeopleFilterOptions for workspace filters and searchPeople for leader filters.

You can select up to 10 survey sends or cycles.

Returned data includes:

• survey context: title, trigger type, anonymity mode, start date, and selected periods
• participation: total recipients, responded recipients, and response rate
• question structure: question text, question type, and available options
• question results: total responses for each question in each selected period
• sentiment: sentiment score where the question supports sentiment analysis
• themes: recurring themes detected from text or sentiment-bearing answers
• option totals: response totals for each selectable answer option

Field notes:

• anonymityMode describes whether responses are anonymous, identity-optional, or identity-required.
• response rate tells you whether there is enough participation to trust the readout.
• sentiment score is a directional signal for how responses trend; read it alongside the question text, response count, and themes.
• themes summarize repeated ideas in text responses. They are best used to identify what to investigate, not as final conclusions by themselves.
• option totals show how many responses selected each option for select-style questions.
• comparing selected periods can show whether participation, sentiment, or answer distribution is changing over time.

Access notes:

• Survey owners can view survey insights and use filters.
• Survey leaders can view insights for their reporting line.
• Users who are neither owners nor leaders cannot view survey insights through MCP.

Useful for engagement dashboards, pulse trend reports, survey theme summaries, and leadership follow-up planning.

Knowledge Search

searchKnowledge

Searches ClarityLoop reference knowledge. Use it when an agent needs product guidance, setup help, workflow instructions, or people-science context before answering or taking action.

This tool does not search workspace people data. It searches ClarityLoop documentation and people-science content.

Inputs include:

• query
• source: docs, science, or both
• result limit up to 10

Returned data includes:

• matching content excerpts
• title and category
• source path
• relevance score
• authors and publication date for people-science results

Use docs for product setup, troubleshooting, and ClarityLoop workflows. Use science for research-backed guidance from ClarityLoop Science on feedback, coaching, growth, performance, and people development. Use both when the question needs product context and people-science context together.

Useful for helping agents answer with ClarityLoop-specific guidance instead of generic advice.

What MCP Does Not Expose

MCP is designed to make ClarityLoop data useful in AI workflows while keeping the same product boundaries people already rely on.

• Personal tools stay personal. For example, feedback given, incoming feedback requests, sent feedback requests, notes, and meetings are centered on the connected user.
• Leadership Insights can find people who match criteria such as sentiment, goal progress, feedback activity, 1:1 activity, completed actions, level, or promotions. Results still follow manager and workspace-owner access.
• Review-cycle data is not exposed through MCP. Feedback, goals, growth signals, work context, and career framework data can support review prep, but review submissions and calibration records are not available.
• Managers do not see feedback that was not shared with them, and anonymous feedback does not reveal the giver's identity.
• Feedback requests are visible to the person who received the request; MCP does not provide a workspace-wide request inbox.
• Notes and meetings are not workspace-wide feeds. They return records visible to the connected user.
• Survey insights are aggregated. Anonymous survey responses should stay anonymous in any dashboard or workflow built on top.
• Sensitive member attributes should keep the same sensitivity in downstream systems.
• Work context is limited to context captured in ClarityLoop and visible to the connected user.
• Write tools create or update real ClarityLoop records as the connected user.
• Store MCP API keys securely and rotate them if exposed.