PresentationMcpServer
Server Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| FREELLMAPI_KEY | No | API key for the free LLM proxy. Optional; defaults to free model routing. | |
| ANTHROPIC_API_KEY | No | API key for Anthropic (premium provider). Required if llm.provider is set to 'anthropic'. |
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
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
| logging | {} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| pingB | Liveness/health check for the Presentation MCP server. Returns 'pong'. |
| ingest_materialsA | Extract text + topic inventory from source files (pdf/docx/pptx/md/txt) within the allowed roots. Image-only PDFs are transcribed via vision. Returns a materialsId used by define_scope/plan_presentation. |
| define_scopeB | Produce the pedagogical Scope Contract: concept graph (core/supporting/stretch/excluded), Bloom targets, prerequisites, boundary statement. Enforces acyclic prerequisites and exclusion reasons in code. |
| plan_presentationB | Build the section outline with per-section time budgets and slide stubs, plus enrichment queries (research gaps). Returns a planId. |
| enrich_planA | Research the plan's knowledge gaps with server-side web search/fetch, within the scope boundary and a search budget. skip:true yields an empty report for offline generation. |
| generate_deckA | The full generation: Draft → Scope-Guard → Render → Script-Sync. Produces a self-contained Reveal.js bundle, presenter script + timing, scope ledger, and a cost-report manifest. Long-running; emits progress. |
| generate_quizzesA | Regenerate misconception-grounded quiz slides on an EXISTING deck and re-render it. Each distractor carries the misconception it targets and a groundedIn provenance pointer. Placed after every N core concepts, before the section summary. |
| render_deckA | Re-render a stored deck from its IR (optionally with a new theme). format 'pdf' is a Phase 3 feature and returns a structured not-available error. |
| rehearse_scriptA | Phase 3 (§11.2). Synthesize each slide's speaker script to audio (Microsoft neural TTS — ONLINE, used at rehearsal time only; the deck bundle stays offline), measure the real durations, and produce a drift report vs the WPM model. If drift exceeds ±10% the timing map is rebalanced to the measured values. The implied effective WPM is stored in the manifest and persisted as the default for the next deck. Emits per-slide mp3s + measured.json + full-run.m3u under the deck's rehearsal/ dir. Requires network; unreachable service returns a structured TTS_UNAVAILABLE error, deck untouched. |
| list_decksA | List all generated decks (id, title, created, duration, slide count). |
| get_deckA | Full manifest + scope-ledger summary + bundle file paths for one deck. |
| delete_deckB | Delete a deck bundle. Refuses unless confirm:true. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 12 tools
Each tool targets a distinct stage in the presentation pipeline (ingestion, scoping, planning, enrichment, generation, rendering, rehearsal, quiz creation, deck management) with no overlapping responsibilities. The CRUD operations for decks are clearly separate from the generation workflow.
Tool names follow a consistent verb_noun snake_case pattern (e.g., list_decks, generate_deck, rehearse_script) with the single exception of `ping`, which is a health check utility and doesn't follow the pattern. This minor deviation doesn't undermine overall consistency.
With 12 tools, the server is well-scoped for its domain—providing a complete end-to-end workflow without redundant or superfluous operations. The count is within the ideal range and each tool serves a clear purpose.
The toolset covers the full presentation lifecycle from material ingestion, pedagogical scoping, planning, enrichment, generation, rendering, rehearsal, and quiz creation to deck management (list/get/delete). No critical operations are missing for the stated purpose.