Skip to main content
Glama

Server Configuration

Describes the environment variables required to run the server.

NameRequiredDescriptionDefault

No arguments

Instructions

Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.

This server publishes no instructions, or was last inspected before Glama recorded them.

Capabilities

Features and capabilities supported by this server

Protocol revision2025-11-25

CapabilityDetails
tools
{
  "listChanged": false
}
experimental
{}

Tools

Functions exposed to the LLM to take actions

NameDescription
quick_conflict_checkA

Check the WCC admin Google Calendar for conflicting events at a proposed time. Run this BEFORE drafting an event with a date. If a conflict is found you MUST STOP and ask the user to pick a different time — never proceed with a conflicting slot. Takes ~10s.

draft_eventA

Phase 1A: turn raw event notes into a structured draft (description, LinkedIn copy) via Gemini, written to /tmp/draft_event.json. Provide as much detail as possible (title, speakers, date/time, format). All event times are interpreted as Europe/London. Takes 30-90s. Only ONE event may be in flight at a time (shared state: /tmp/draft_event.json). Use cleanup_state to abandon a draft and start over.

speaker_lookupB

Search the WCC Google Drive speaker database. action is one of: search | download | create | upload_headshot | upload_bio. 'search "Jane Doe"' finds a profile; 'download' fetches the headshot. Use before drafting to reuse existing speaker bios. Takes ~10s.

set_registration_linkA

Save the registration URL (Meetup/Luma/Google Form) into the current event draft. REQUIRED before publish_event — publishing without a registration link is forbidden. Instant.

create_registration_formA

Create a public Google registration Form for the event and store its URL in the draft. Use when the organizer has no Meetup/Luma link. Takes ~30s.

publish_eventA

Phase 2: publish the drafted event — creates the Drive folder, uploads event-details.json, creates the Google Calendar event with Meet link, creates the Luma listing, and queues social posts. PRECONDITIONS: a draft exists AND set_registration_link has succeeded. Meetup listing and poster generation are NOT automated in this version. Takes several minutes — wait for the result. Only ONE event may be in flight at a time (shared state: /tmp/draft_event.json). Use cleanup_state to abandon a draft and start over.

schedule_linkedinA

Generate the 11-post LinkedIn cadence timestamps for an event date (10 days out through event day, 10:00 Europe/London). Returns a JSON array of ISO timestamps. Instant.

collect_attendanceA

Phase 3: pull attendee lists from Luma and Meetup and append them to the master Google Sheet. Takes ~1-2 minutes.

process_recordingA

Phase 3: find the event recording + transcript on Drive, trim silence, upload to YouTube, and write links back into the event's event-details.json. VERY SLOW — can take 15-40 minutes; wait for the result and do not retry while it runs. Only ONE event may be in flight at a time (shared state: /tmp/draft_event.json). Use cleanup_state to abandon a draft and start over.

cleanup_stateA

Delete the in-flight draft state (/tmp/draft_event.json and poster temp files). Use to abandon a draft or after a completed publish. Pass force=true to ignore the 30-minute idle guard.

Prompts

Interactive templates invoked by user choice

NameDescription

No prompts

Resources

Contextual data attached and managed by the client

NameDescription

No resources

TDQS

A4.1/5.0

Scored across 10 tools

Disambiguation5/5

Each tool targets a distinct stage in the event lifecycle: pre-check, drafting, speaker lookup, registration, scheduling, publishing, post-event processing, and cleanup. Even set_registration_link and create_registration_form are clearly separated (one saves an existing link, the other creates a new form). No two tools are ambiguous.

Naming Consistency4/5

Most tools follow a verb_noun pattern (draft_event, publish_event, collect_attendance, cleanup_state). Minor deviations exist: quick_conflict_check uses a modifier + noun_verb, and speaker_lookup has object_verb order. Overall the pattern is predictable and readable.

Tool Count5/5

10 tools is well within the ideal range and each tool serves a necessary step in the event management workflow. The count feels intentionally scoped, covering planning through post-event actions without unnecessary bloat.

Completeness4/5

The surface covers the full event lifecycle: conflict checking, drafting, speaker lookup, registration, publishing, attendance, recording, and cleanup. Minor gaps exist (e.g., no update/cancel for published events, poster generation noted as not automated), but the core workflow is complete and agents can execute end-to-end.

Maintenance

ActivityStale
ResponsivenessNo issues