fvtt-mcp-molten5e: Skill for Claude Code

.claude/skills/session-scribe/SKILL.md

session-scribe is a skill for Claude Code from Txpple/fvtt-mcp-molten5e. It costs 204 tokens per session (4,720 once invoked), scanned A, original, MIT.

A workflow for turning a Craig recording and a Foundry chat log into a speaker-labeled transcript and session notes. Foundry is a virtual tabletop for role-playing games.

In plain words
What is it for?
Use it after a game session to create a recap, game-master notes, and a player-safe HTML recap in the campaign repository.
Why use it?
It aligns spoken audio with game events such as rolls, whispers, and cards without asking the user to add markers manually.

Skill for Claude Code

Written for Claude Code: installed under .claude/. Also seen: mentions Claude Code.

This is Txpple/fvtt-mcp-molten5e's own configuration. It tells Claude Code how to work on fvtt-mcp-molten5e itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything fvtt-mcp-molten5e configures →

Reuse

Borrowing it

Nothing to install: this file belongs to Txpple/fvtt-mcp-molten5e. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/Txpple/fvtt-mcp-molten5e/main/.claude/skills/session-scribe/SKILL.md
Clone the repo
git clone --depth 1 https://github.com/Txpple/fvtt-mcp-molten5e

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for session-scribe

README.md
[![agentmods](https://agentmods.dev/badge/skills/txpple/fvtt-mcp-molten5e/session-scribe/github.svg)](https://agentmods.dev/skills/txpple/fvtt-mcp-molten5e/session-scribe)
Your own site
<a href="https://agentmods.dev/skills/txpple/fvtt-mcp-molten5e/session-scribe"><img src="https://agentmods.dev/badge/skills/txpple/fvtt-mcp-molten5e/session-scribe/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for session-scribe

Your own site · 80×15
<a href="https://agentmods.dev/skills/txpple/fvtt-mcp-molten5e/session-scribe"><img src="https://agentmods.dev/badge/skills/txpple/fvtt-mcp-molten5e/session-scribe.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 204 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 4,720 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00204 $0.04720
Opus 5 $0.00102 $0.02360
Sonnet 5 $0.00041 $0.00944
Haiku 4.5 $0.00020 $0.00472

Measured 5d ago against content hash e49a50e350d1, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

session-scribe scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

The scan reads SKILL.md. This mod also ships 2 executable files (scripts/session_scribe.py, scripts/setup.ps1), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

.claude/skills/session-scribe/SKILL.md · 237 lines

How it starts

The opening of the file, as written. The whole thing — 237 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Session scribe

Craig records the table (one audio track per speaker); Foundry's chat log records the mechanics (every roll, whisper, and card, each with an epoch-ms timestamp). Craig's recording metadata carries its own startTime — so the two timelines align by pure wall-clock arithmetic. No sync ritual, no markers, no Discord integration. The user pastes one link; everything else is yours.

The user's contract (locked 2026-07-06, notes-repo memory session-recording-pipeline): /join Craig at session start → play → /stop → paste Claude the link. That is ALL. Do not ask them to mark, note, export, or download anything. If they said "mark that" aloud during play, it is IN the transcript — grep for it.

Machine prerequisites (once per machine)

Run scripts/setup.ps1 (idempotent — installs ffmpeg + uv via winget, builds the ~\.session-scribe\venv with faster-whisper + CUDA wheels, smoke-tests the GPU stack):

powershell -ExecutionPolicy Bypass -File .claude\skills\session-scribe\scripts\setup.ps1 -PrefetchModel

SMOKE TEST OK = CUDA works. OK (CPU ONLY) = transcription still works, just slower — on a new GPU generation this usually means ctranslate2 needs a version bump for the new compute capability (uv pip install --python ~\.session-scribe\venv\Scripts\python.exe -U ctranslate2), then re-run the smoke test. Machines verified: RTX 4070 laptop ✅ (2026-07-06), DESKTOP-NY ✅ (2026-08-25, cuda/float16 — ~2.9h × 5 tracks in <5 min, on signed Python 3.13; hardened setup.ps1 re-verified end-to-end there 2026-08-26, incl. the signed-pick venv-creation path).

Known failure — Windows Application Control blocks the uv-managed Python (hit on DESKTOP-NY 2026-08-25): ImportError: DLL load failed while importing _ctypes: An Application Control policy has blocked this file. The uv/python-build-standalone interpreter's DLLs trip the policy; a venv on a SIGNED python.org install passes. Fix: winget install Python.Python.3.13 → delete ~\.session-scribe\venv → re-run setup.ps1 (it prefers the newest signed 3.12+ install when one exists, and its ctypes probe fails fast with these instructions BEFORE the ~1.3 GB wheel download — the smoke test would also catch the block, but only after it). faster-whisper + CUDA wheels work fine on 3.13. Never work around this by touching the App Control policy itself. If a broken venv was renamed aside as ~\.session-scribe\venv-blocked-* (~1.3 GB each), delete those backups once the rebuilt venv passes the smoke test — setup.ps1 lists any it finds.

Read the full file on GitHub · 237 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago Changed · +31 lines e49a50e350d1
  2. 9d ago First seen · 206 lines · 204 tokens per session scan A 31f17778d973

Subscribe to this mod's changes

session-scribe is a skill published in the GitHub repository Txpple/fvtt-mcp-molten5e (0 stars, last pushed 8d ago), licensed MIT. It adds 204 tokens to every session and 4,720 once invoked, about $0.0010 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.