direct-a-voice-recording

Direct a safe, consistent voice-recording session with performance context, technical checks, consent, and usable coverage. Use when recording narration, dialogue, prompts, or voiceover.

Jump to install

Repository: protosphinx/sphinxstack
Last source activity: July 29, 2026 at 13:55
Detected SKILL.md language: English
Stars: 1
Forks: 0

Install options

The review-first prompt is selected by default. You can switch to a direct command or download a local copy.

Install with Codex or Claude Copy this prompt, paste it into Codex, Claude, or another assistant, and let it review the skill page and install it for you.

A direct command skips the review prompt. Inspect the source before running it.

npx skills add https://github.com/protosphinx/sphinxstack --skill direct-a-voice-recording

The command stays on one line. Scroll horizontally to inspect it before copying.

Prefer a local copy? Download the files currently available to SkillsMP.

Review the source files

Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.

File Explorer

2 files

SKILL.md

Source instructions · Read-only preview

Back to search Back to installation

assign-material-and-check-mass-in-partmode

Assign materials to PartMode bodies and verify revision-bound volume, mass, center of mass, inertia, and geometry health. Use when someone asks how much a part or assembly weighs, needs density-driven properties, or sees an incomplete mass result.

2026-08-08

connect-partmode-to-an-agent

Configure and verify PartMode's hosted MCP server for a person's agent without exposing credentials or granting more CAD authority than needed. Use when they say "connect PartMode to Codex," "set up PartMode MCP," "add PartMode to my agent," or ask why the PartMode MCP URL does not behave like a normal web page.

2026-08-08

create-a-drawing-in-partmode

Generate and verify an exact projected PartMode drawing from a settled part, assembly, or configuration. Use when someone asks for a drawing sheet, SVG, DXF, or PDF output, standard views, dimensions, notes, or a reviewable manufacturing document.

2026-08-08

View all from this repository

direct-a-voice-recording

Create conditions for a repeatable performance without pushing the speaker past comfort, consent, or vocal safety.

Procedure

Confirm script version, audience, tone, pronunciation, deliverables, usage rights, session length, and approval owner.

Review consent, compensation, identity or synthetic-voice restrictions, confidentiality, retakes, and future-use boundaries.

Mark the script for meaning, emphasis, pace, pauses, pronunciation, alternate readings, and pickups.

Prepare a quiet, accessible space with water, breaks, communication, and safe microphone placement.

Set recording format, sample rate, bit depth, input, distance, gain, monitoring, file naming, and redundant capture where needed.

Record room tone and a test; listen for clipping, noise, reflections, plosives, mouth noise, latency, and interference.

Give direction in terms of audience, intention, and specific observable adjustment rather than personal judgment.

Capture complete takes, targeted pickups, difficult names, alternate pacing, and intentional silence.

Stop for strain, discomfort, changed consent, technical failure, or unsafe conditions.

Verify files, notes, backups, selected takes, rights records, and handoff before release.

Failure plan

Never clone, transform, or reuse a voice beyond documented consent.

Do not pressure a performer through pain, distress, identity-sensitive material, or revoked consent.

If audio integrity is uncertain, pause and monitor a fresh test rather than continuing unusable capture.

Restrict unreleased scripts and raw voice files appropriately.

Done

Consent, script, pronunciations, technical format, and usage are confirmed

Test, room tone, full takes, pickups, and two verified copies are recorded

Session notes and rights evidence travel with the accepted files