Mantella Config

---
name: mantella-config
description: Edit, merge, or diagnose Mantella configuration, prompts, conversation flow, TTS integration, and source changes
  while preserving user settings and secrets.
compatibility: Windows 10/11; Skyrim Special Edition or Anniversary Edition; PowerShell and Python 3 when bundled scripts
  are used
metadata:
  version: 4.3.0
  updated: '2026-07-22'
  library: overseer-skyrim-agent-skills
  provider: claude
  provider_pack_version: 1.0.0
  base_library: Skyrim-Agent-Skills-v6
  error_registry_revision: 4.3.0
when_to_use: Use for edit, merge, or diagnose mantella configuration, prompts, conversation flow, tts integration, and source
  changes while preserving user settings and secrets.
---

# Mantella configuration and source

## Inputs

- Exact Mantella version and installation type.
- Current config, prior config, logs, source revision, and desired behavior.
- LLM and TTS provider contract without exposing secrets.

## Workflow

1. Back up the current config and redact credentials from reports.
2. Diff old and new schemas by key, section, default, type, and deprecation.
3. Preserve user values only where the new schema still supports them.
4. Keep prompt changes separate from transport, TTS, interruption, and Papyrus changes.
5. Validate config parsing and start the narrowest component possible.
6. For source edits, map the conversation lifecycle and cancellation/interruption paths before changing code.
7. Test start, normal response, interruption, timeout, TTS failure, provider failure, and shutdown.
8. Document any setting that must be re-entered manually.

## Context efficiency

Do not load entire voice-model folders or conversation histories. Use the failing config, relevant source modules, and bounded logs.

## Evidence standard

Use this hierarchy for version-sensitive facts:

1. The active project's `ENVIRONMENT.md`, `TASK.md`, installed framework version, and build files.
2. Official upstream documentation or source for that exact version.
3. Known-good files from the user's installed mod library, read-only.
4. Direct inspection of the relevant ESM, ESP, DLL, script, log, or archive.

Do not substitute memory, an old example, or a plausible token. Record the evidence path or URL in `VALIDATION.md`.

## Claude Code execution adapter

- Keep `CLAUDE.md` concise and use this skill for procedural detail.
- Load only the skills needed for the current milestone because invoked skill content remains in context.
- Use high or xhigh effort for risky architecture, plugins, DLLs, and hostile review.
- Before compaction or a usage boundary, persist exact state, commands, uncommitted changes, and remaining validation.
- Treat auto memory as candidate learning, not verified truth, until it passes the memory-promotion protocol.

### Skill-specific provider control

Preserve user secrets and local settings; separate configuration compatibility from source-code changes.