Skip to main content
Glama

Server Details

Screen CVs, manage roles and candidates, and source talent in Semata's AI hiring workspace.

Ownership verified
Status
Healthy
Uptime
100.0% over the last 24 hours
OAuth
Works in Glama
Last Tested
Transport
Streamable HTTP · MCP 2025-11-25
URL
ChatGPT
–
Claude
–

TDQS

A3.6/5.0

Scored across 54 tools

Disambiguation4/5

Despite 54 tools, descriptions actively carve out boundaries: reject_candidate vs set_candidate_rating(gated) vs bulk_triage_candidates vs record_hiring_outcome are explicitly distinguished, as are the three report tools and the sourcing build/edit/run trio. The residual risk is on near-neighbour creation paths (create_role vs create_role_from_jd vs parse_job_description vs duplicate_role) and the various review/inbox tools, which are similar enough to occasionally misselect.

Naming Consistency5/5

Uniform snake_case with verb-first naming across the entire set: add_*, list_*, get_*, create_*, update_*, set_*, record_*, remove_*, request_*, rerun_*, restore_*, reopen_*, resolve_*, build_*, edit_*, source_*, import_*, invite_*. No camelCase or noun-only outliers breaking the pattern.

Tool Count2/5

54 tools is well past the comfortable range and forces agents to navigate a large surface for a single request. The breadth is partly justified by covering sourcing, screening, reviews, interviews, outcomes, ATS import and reporting, but there is clear consolidation opportunity (e.g. rating/reject/triage/stage verbs, and the parallel get_*_report family) that would shrink the set materially.

Completeness4/5

Coverage is unusually thorough: full role lifecycle (create from text/JD, duplicate, close, criteria update/restore), stage management, sourcing pipeline, candidate review and hiring outcome recording with reopen, interview question sets, ATS import and multiple reports. Minor gaps remain, such as no role-level metadata update beyond criteria, no role reopen/unarchive, and no deletion paths for comments or candidates.

Available Tools

54 tools
add_candidate_commentAdd candidate commentAInspect

Add a comment to a candidate, or reply into an existing root thread by passing parentId. Viewers may add comments.

Inputs: content (required), candidateId (required), workspaceId (required), parentId.

add_role_noteAdd role noteAInspect

Add a note of up to 5000 characters to a role, shown to everyone with access to the role and attributed to the signed-in user. Adds a new note and changes nothing that exists; call it again to add another, as notes cannot be edited or removed here. Viewers may add notes. Requires OAuth or a scoped token with collaboration access.

Inputs: roleId (required), content (required), workspaceId (required).

add_sourced_candidates_to_screeningAdd sourced candidates to screeningAInspect

Add people from a source_candidates result to a role's screening and run them through the CV pipeline. This consumes the user's daily processing limit; people already in the screening are skipped. Pass either the role id or the screening id, not both. The screening must be active. Requires OAuth and a workspace role that can edit.

Inputs: workspaceId (required), sourcedCandidateIds (required), roleId, batchId.

build_sourcing_queryBuild sourcing queryA
Read-only
Inspect

Build a reviewable sourcing query from a natural-language brief and, optionally, a role's criteria. Always build a query before sourcing and show it to the user for review. This does not search for people. Resolve named employers with resolve_company first. To revise a reviewed query, use edit_sourcing_query instead of building again, unless the role, the brief or the kind of person wanted has changed. Use strategy archetype for people defined by what they did or were, such as founders, former military officers or career changers, rather than by employer. Requires OAuth.

Inputs: workspaceId (required), q, cities, roleId, +14 more.

bulk_triage_candidatesBulk triage candidatesA
Destructive
Inspect

Apply one rating or screening-gate (screen out) action to up to 100 candidates in one screening. This is not a hiring decision; use record_hiring_outcome to record hired, rejected or withdrawn.

Inputs: action (required), batchId (required), workspaceId (required), candidateIds (required), +1 more.

close_roleClose roleA
Destructive
Inspect

Close (archive) a role. Only the role status changes: its screening, candidates, notes and criteria are kept, and a role that is already closed is left as it is. A closed role no longer accepts applications from its public roles page. Requires OAuth or a scoped token and a workspace role that can edit.

Inputs: roleId (required), workspaceId (required).

create_roleCreate roleAInspect

Create a role and its empty screening in the selected workspace. Needs a name and a seniority range in years (0 to 20, minimum not above maximum); a description, target archetypes, up to 7 key requirements, industry experience, company stage and a location are optional. Adds a new record and changes nothing that exists. Retrying the same request returns the role it already created. Requires OAuth or a scoped token and a workspace role that can edit.

Inputs: name (required), workspaceId (required), seniority_max (required), seniority_min (required), +7 more.

create_role_from_jdCreate role from job descriptionAInspect

Create a role and its empty screening from the original pasted text, PDF, or public job-posting URL. The complete original is saved separately from the role summary and screening criteria. Pass exactly one source. Requires OAuth or a scoped token, and workspace write permission.

Inputs: workspaceId (required), url, text, filename, +1 more.

create_role_stageCreate role stageAInspect

Add a custom hiring stage to the end of a role board. Unreviewed and Archived are reserved names, and stage names are unique per role. Requires OAuth and a workspace role that can edit.

Inputs: name (required), roleId (required), workspaceId (required).

delete_role_stageDelete role stageA
Destructive
Inspect

Delete an empty custom stage from a role, by its name from list_role_stages. Unreviewed and Archived cannot be deleted. A stage that still holds candidates is refused with blockingCandidates; move them with set_candidate_stage first. Requires OAuth and a workspace role that can edit.

Inputs: stage (required), roleId (required), workspaceId (required).

duplicate_roleDuplicate roleAInspect

Copy a role, its description and its criteria into a new role with its own empty screening. Candidates, notes and comments are not copied, and the source role is unchanged. The copy is active even when the source is closed, and is named " (copy)" unless name is given (up to 200 characters). Retrying the same request returns the copy it already made. Requires OAuth or a scoped token and a workspace role that can edit.

Inputs: roleId (required), workspaceId (required), name.

edit_sourcing_queryEdit sourcing queryA
Read-only
Inspect

Apply one clear recruiter change to the latest reviewed sourcing query, keeping unaffected clauses. Returns the edited query and feedback whose status is possible, partial, or impossible. This does not search for people. Show the edited query to the user before sourcing it. Resolve any newly named employer with resolve_company and include its name and LinkedIn URL in editRequest. Requires OAuth.

Inputs: pdlQuery (required), editRequest (required), workspaceId (required).

get_ats_connectionsGet ATS connectionsA
Read-only
Inspect

Check which ATS providers are connected to the selected workspace.

Inputs: workspaceId (required).

get_batchGet screeningA
Read-only
Inspect

Get one screening by its id, including its processing counts and available candidate summary data. Resolve a screening name with the screening list tool before calling this tool.

Inputs: batchId (required), workspaceId (required), jobLimit, jobOffset, +1 more. Returns: code, jobs, batch, error, +1 more

get_candidateGet candidateA
Read-only
Inspect

Fetch one candidate summary from the selected Semata workspace. Raw CV content and internal model inputs are not returned.

Inputs: candidateId (required), workspaceId (required). Returns: code, batch, error, candidate, +1 more

get_candidate_reportGet candidate reportA
Read-only
Inspect

Get the content of a candidate's scorecard report: ranking, experience, strengths, considerations, criteria verdicts, skills, gaps and suggested interview focus areas. Semata renders the PDF in the browser, so this returns the report as data plus appUrl, the page where the user can download the PDF. Raw CV text is not returned.

Inputs: candidateId (required), workspaceId (required).

get_hiring_reportGet hiring reportA
Read-only
Inspect

Outcome-based hiring report for the selected workspace, computed from recorded decisions and activity events, never from scores or the screening gate. cohort = applications created in the period by current outcome; decisions = outcome events recorded in the period. Funnel, stage age and time to hire use recorded history only and are marked unavailable or partial when it is missing; applications created before outcome tracking began are legacy and excluded from funnel and stage age. Rates are withheld for small cohorts.

Inputs: workspaceId (required), to, from, roleIds, +2 more.

get_job_statusGet candidate job statusB
Read-only
Inspect

Get bounded processing status and a safe failure reason for one candidate.

Inputs: candidateId (required), workspaceId (required).

get_review_queueGet role review queueA
Read-only
Inspect

List active evaluated candidates for a role, your verdicts, team verdict counts and your review progress. Defaults to candidates you have not reviewed. Disagreement means a yes or superstar and a no. Requires OAuth or a scoped token.

Inputs: roleId (required), workspaceId (required), limit, offset, +2 more.

get_roleGet roleA
Read-only
Inspect

Get one role by its id, with its defined criteria and its screening. Resolve a role name with the role list tool before calling this tool.

Inputs: roleId (required), workspaceId (required). Returns: code, role, batch, error, +1 more

get_saved_interview_questionsRead saved interview questionsA
Read-only
Inspect

Read a prepared set without generating again. Omit planId for the latest set or pass an id from list_interview_question_sets. Changed evidence is marked stale. Requires OAuth or a scoped token.

Inputs: candidateId (required), workspaceId (required), planId.

get_screening_reportGet screening reportA
Read-only
Inspect

Get the content of a screening report: the role criteria, processing counts, and evaluated candidates ranked by score (up to 100 per page). Semata renders the PDF in the browser, so this returns the report as data plus appUrl, the page where the user can download the PDF.

Inputs: batchId (required), workspaceId (required), limit, offset.

import_ats_candidatesImport ATS candidatesAInspect

Import 1 to 5 candidates from a connected ATS (workable, pinpoint, greenhouse or ashby) into a Semata screening, and queue their CVs for processing. Use list_ats_candidates for the candidate ids. Give destinationBatchId for an existing screening or destinationRoleId for a role's screening, not both. A closed screening is refused. Candidates already imported are skipped and reported, each import counts against the daily CV processing limit, and the call fails if the limit would be exceeded. Each candidate succeeds or fails on its own: read every entry in results. Requires OAuth or a scoped token with ATS access and a workspace role that can edit.

Inputs: provider (required), workspaceId (required), candidateIds (required), destinationRoleId, +1 more.

invite_organisation_memberInvite organisation memberA
Destructive
Inspect

Invite a person by email to an organisation workspace as a viewer, member (the default) or admin; owners cannot be invited. Creates a pending invitation and sends an email to that address, which cannot be recalled by this server. Only organisation admins and owners can call it, and personal workspaces are refused. Fails when the person is already a member or already has a pending invitation. If the email cannot be sent the invitation still exists and has to be revoked in Semata before trying again. Requires OAuth or a scoped token with invite access.

Inputs: email (required), workspaceId (required), role.

list_ats_candidatesList ATS candidatesC
Read-only
Inspect

List bounded candidates from one ATS job.

Inputs: jobId (required), provider (required), workspaceId (required).

list_ats_jobsList ATS jobsC
Read-only
Inspect

List bounded open jobs from one connected ATS provider.

Inputs: provider (required), workspaceId (required).

list_batchesList screeningsA
Read-only
Inspect

List screenings (candidate screening lists) in the selected Semata workspace. Every role has exactly one screening, so filtering by role id returns at most one. Use this to resolve a screening name to its id before calling the screening detail tool.

Inputs: workspaceId (required), limit, offset, roleId, +1 more. Returns: code, error, limit, total, +3 more

list_candidate_commentsList candidate commentsA
Read-only
Inspect

List up to 100 comments and candidate activity events, newest first. Root threads have parentId null; replies carry the root comment id.

Inputs: candidateId (required), workspaceId (required).

list_candidate_reviewsList candidate reviewsA
Read-only
Inspect

List teammates' verdicts (superstar, yes, or no) and reasons on a candidate, most recently updated first. Paginated; defaults to 25 reviews.

Inputs: candidateId (required), workspaceId (required), limit, offset.

list_candidatesList candidatesA
Read-only
Inspect

List bounded candidate summaries in a selected workspace screening. Resolve the screening with list_batches first.

Inputs: batchId (required), workspaceId (required), limit, offset, +1 more.

list_hiring_outcome_reasonsList hiring outcome reasonsA
Read-only
Inspect

List the reasons that can be recorded with a hiring outcome: the standard reasons plus the selected workspace's own. Rejected and withdrawn outcomes require one; hired does not.

Inputs: workspaceId (required), outcome.

list_interview_question_setsList saved interview questionsA
Read-only
Inspect

List the ten most recent prepared sets for this candidate application. A set does not prove an interview happened. Requires OAuth or a scoped token.

Inputs: candidateId (required), workspaceId (required).

list_reviewersList reviewersA
Read-only
Inspect

List teammates in this workspace who can be asked to review a candidate, with their user ids. Requires OAuth or a scoped token.

Inputs: workspaceId (required).

list_review_inboxList your review inboxA
Read-only
Inspect

Read your review inbox in this workspace: toReview holds candidates a teammate sent you for review that you have not given a verdict on; updates holds mentions, replies, and verdicts on candidates you asked about or took part in. Filter by section or unread only. Reading does not mark anything read. Requires OAuth or a scoped token.

Inputs: workspaceId (required), limit, offset, section, +1 more.

list_review_rolesList roles to reviewA
Read-only
Inspect

List open roles in this workspace with candidate counts and your review progress. Requires OAuth or a scoped token.

Inputs: workspaceId (required), limit, offset.

list_role_notesList role notesB
Read-only
Inspect

List up to 100 notes on a role, newest first.

Inputs: roleId (required), workspaceId (required).

list_rolesList rolesA
Read-only
Inspect

List the roles in the selected Semata workspace. Each role represents one hiring need and owns exactly one screening list of candidates. Call this to see what the user is hiring for, or to resolve a role name to its id before calling the role detail tool.

Inputs: workspaceId (required), limit, offset, includeClosed. Returns: code, error, limit, roles, +3 more

list_role_stagesList role stagesA
Read-only
Inspect

List a role's hiring stages in board order. Unreviewed and Archived are always available. Use the returned names when moving candidates with set_candidate_stage or deleting a stage.

Inputs: roleId (required), workspaceId (required).

list_workspacesList workspacesA
Read-only
Inspect

List the personal and organisation workspaces available to the authenticated Semata user. Call this before any workspace-scoped tool and pass the selected workspace id to every subsequent call.

Inputs: no named parameters. Returns: code, error, workspaces

parse_job_descriptionParse job descriptionA
Read-only
Inspect

Turn a job description into a role draft for the user to review before creating the role. Pass either the pasted text or a public job-posting URL, not both; Semata fetches the URL itself with its safe fetcher. Nothing is saved. After review, use create_role_from_jd with the original text or URL to create the role and retain its full original description. create_role accepts a draft but has no original source. Requires OAuth or a scoped token.

Inputs: workspaceId (required), url, text.

prepare_interview_questionsPrepare interview questionsAInspect

Prepare and save interview questions by candidate application and round, using the CV, role criteria, and current earlier interview evidence. Prepared sets are plans, not proof questions were asked. Pass round when named, focus for time limits or topics, and includeRoleCriteria false for a role-agnostic set. Requires OAuth or a scoped token.

Inputs: candidateId (required), workspaceId (required), focus, round, +1 more.

record_hiring_outcomeRecord hiring outcomeA
Destructive
Inspect

Record that up to 100 applications in one screening were hired, rejected or withdrawn. This is the final human decision, separate from the screening gate, stage, archiving and reviewer verdicts, and it never changes scores or evidence. Rejected and withdrawn need a reasonId from list_hiring_outcome_reasons. An application that already has a different outcome must be reopened first. Each application succeeds or fails on its own: read every entry in results.

Inputs: batchId (required), outcome (required), workspaceId (required), candidateIds (required), +2 more.

reject_candidateReject candidateA
Destructive
Inspect

Screen out one completed candidate: apply the screening gate so they leave the active shortlist (processing status rejected). This is not a hiring decision and records no hiring outcome; use record_hiring_outcome for that.

Inputs: candidateId (required), workspaceId (required).

remove_candidate_reviewRemove your candidate reviewA
Destructive
Inspect

Remove only the signed-in user's verdict and reason from a candidate. Does not change workflow stage or teammates' reviews. Viewers may remove their own review. Requires OAuth or a scoped token.

Inputs: candidateId (required), workspaceId (required).

reopen_hiring_outcomeReopen hiring outcomeA
Destructive
Inspect

Reopen up to 100 applications that have a hired, rejected or withdrawn outcome so they are active again. The earlier decision stays in the history; stage, archive state and screening evidence do not change. Each application succeeds or fails on its own: read every entry in results.

Inputs: batchId (required), workspaceId (required), candidateIds (required), note.

request_candidate_reviewSend candidate for reviewA
Destructive
Inspect

Ask named teammates for a verdict on a candidate, with an optional message. Each reviewer is notified and the candidate lands in their To review inbox until they post a verdict. Resolve reviewer ids with list_reviewers first. Requires OAuth or a scoped token.

Inputs: candidateId (required), reviewerIds (required), workspaceId (required), message.

rerun_criteria_matchRerun criteria matchA
Destructive
Inspect

Queue criteria matching again for the candidates in a screening, so their criteria scores are recalculated against the role as it is now. Only completed, scored candidates that are not archived are eligible; others are skipped and counted. Each queued candidate counts against the daily CV processing limit, and the call fails without queuing anything when the limit would be exceeded. A closed screening is refused. Re-scoring runs in the background: the result reports how many were queued and skipped, and candidates keep their earlier scores until the new ones are ready. Requires OAuth or a scoped token and a workspace role that can edit.

Inputs: batchId (required), workspaceId (required).

resolve_companyResolve companyA
Read-only
Inspect

Identify a named employer before using it as a sourcing filter. Pass the company name and any LinkedIn company URL or website the user supplied or that you verified on the web; never invent URLs. Checks PDL's company directory only, does not search for people, and says nothing about how many candidates exist. A supplied LinkedIn URL stays usable even when the directory cannot match it; source and providerStatus tell these outcomes apart. Pass the returned name and slug to build_sourcing_query, or mention the name and LinkedIn URL in an edit_sourcing_query request. Requires OAuth.

Inputs: workspaceId (required), name, website, linkedinUrl.

restore_role_criteriaRestore role criteriaA
Destructive
Inspect

Undo the most recent update_role_criteria or restore_role_criteria change on a role. Only use this when the user asks to undo. Pass previousCriteria and previousDescription from that result unchanged; never reconstruct them. Eligible existing candidates are brought in line with the restored criteria. Requires OAuth or a scoped token, and a workspace role that can edit.

Inputs: roleId (required), workspaceId (required), previousCriteria (required), previousDescription (required).

set_candidate_ratingSet candidate ratingA
Destructive
Inspect

Set the recruiter rating on one candidate: favourite, approved, review or gated, replacing any earlier rating. Rating a candidate gated also archives them, so they leave the active shortlist; the other ratings leave the candidate where they are. Use bulk_triage_candidates to rate several at once. This is advice for the team, not a hiring decision; use record_hiring_outcome to record one. Requires OAuth or a scoped token and a workspace role that can edit.

Inputs: rating (required), candidateId (required), workspaceId (required).

set_candidate_reviewSet candidate reviewA
Destructive
Inspect

Record or replace the signed-in user's verdict on a candidate (superstar, yes, or no) with an optional reason. A verdict is an opinion and does not move the candidate between workflow steps. Viewers may review.

Inputs: verdict (required), candidateId (required), workspaceId (required), comment.

set_candidate_stageSet candidate stageA
Destructive
Inspect

Move up to 100 candidates in one screening to a named stage of its role. Use a stage name from list_role_stages; pass Unreviewed (or null) to clear the stage and Archived to archive the candidates. Candidates outside the screening are skipped and counted. Requires OAuth and a workspace role that can edit.

Inputs: stage (required), batchId (required), workspaceId (required), candidateIds (required).

source_candidatesSource candidatesAInspect

Run a reviewed sourcing query once and save the matching people to Semata's sourced candidate pool. It returns at most 20 people per call. Only call it after the user has reviewed the query from build_sourcing_query or edit_sourcing_query and explicitly asked to source it. Default to 10 people per call. 20 is the maximum per call, so a larger request takes several narrower searches. Pass the query unchanged. People you already sourced for the same workspace and role are excluded, and every request is audited. Returns candidate summaries, a total, and a status of matches or no_match; a failure is an error, never an empty result. Add chosen candidates to a role with add_sourced_candidates_to_screening. Requires OAuth and a workspace role that can edit.

Inputs: pdlQuery (required), workspaceId (required), limit, roleId, +1 more.

update_role_criteriaUpdate role criteriaA
Destructive
Inspect

Update only the supplied criteria fields on a role. Eligible existing candidates are brought in line with the new criteria. The result includes previousCriteria and previousDescription; pass them unchanged to restore_role_criteria to undo the change.

Inputs: roleId (required), workspaceId (required), setup, languages, +15 more.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    Related MCP Connectors

    Related MCP Servers

    Try in Browser

    Glama MCP Gateway

    Add one secure layer between your agents and this server.

    Resources