A voice-control skill for interrupting an active Cadence Code conversation and adding a spoken instruction to the current coding task. Cadence Code is a local voice session used with Codex or Antigravity.

Use it when you explicitly invoke the skill to stop the current audio turn and capture one new instruction. It can also report when no speech or no active session is available.

It lets you redirect a running task without ending the voice session or waiting for its current turn to finish.

jump-in — Skill for Claude Code and Codex by michael-L-i

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Any agent

npx skills add michael-L-i/cadence-code --skill jump-in

Clone the repo

git clone --depth 1 https://github.com/michael-L-i/cadence-code

Made for: Claude Code, Codex.

Or install cadence-code, the plugin that ships this one along with the rest of its 4 skills, 4 commands.

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.

README.md

[![agentmods](https://agentmods.dev/badge/skills/michael-l-i/cadence-code/jump-in.svg)](https://agentmods.dev/skills/michael-l-i/cadence-code/jump-in)

Your own site

<a href="https://agentmods.dev/skills/michael-l-i/cadence-code/jump-in"><img src="https://agentmods.dev/badge/skills/michael-l-i/cadence-code/jump-in.svg" alt="Measured on agentmods" height="20"></a>

Per session 47 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.

When invoked 422 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.

Model	Per session	Once invoked
Fable 5.1	$0.00047	$0.00422
Opus 5	$0.00023	$0.00211
Sonnet 5	$0.00009	$0.00084
Haiku 4.5	$0.00005	$0.00042

Measured 7d ago against content hash bcb7000eb880, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-07, from the pricing page.

Security

Grade A, and why

jump-in 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 7d ago.

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.

skills/jump-in/SKILL.md · 33 lines

What it actually says

Jump In

Use mcp__cadence-code__voice_interrupt to stop current Cadence Code audio and capture one fresh spoken instruction. Cadence Code keeps its local speech models loaded; this does not end the Cadence Code session. Antigravity may display the tool as voice_interrupt under the cadence-code MCP server.

Call mcp__cadence-code__voice_interrupt immediately. Do not call voice_start, voice_speak, or ordinary voice_listen first.
If it returns ok: false, show the error. When there is no active session, tell the user to invoke $start-talking in Codex or /start-talking in Antigravity; do not start Cadence Code implicitly.
Treat every non-empty transcript as added guidance for the interrupted task, including one returned with end_reason: "timeout". Continue from the current conversation and repository state without asking whether to resume. Follow the guidance normally when it redirects or replaces the earlier work.
If no speech is detected, leave the interrupted task stopped and report that no guidance was captured. Do not listen again automatically.
On end_reason: "device_error", show the microphone problem and leave the Cadence Code session active so the user can retry or invoke $wrap-up or /wrap-up.
If the transcript clearly asks to end Cadence Code entirely, call mcp__cadence-code__voice_stop and do not resume the task.

After accepting added guidance, continue the normal Start Talking behavior: work silently, then provide the concise spoken result and detailed written result before listening for the next turn.

Files

What ships with it

1 file 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.

agents/openai.yaml 218 B

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.

7d ago First seen · 33 lines · 47 tokens per session scan A bcb7000eb880

Subscribe to this mod's changes

jump-in is a skill published in the GitHub repository michael-L-i/cadence-code (2 stars, last pushed 6d ago), licensed MIT. It adds 47 tokens to every session and 422 once invoked, about $0.0002 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.

Apple Silicon Claude Code claude-code-plugin mlx speech-to-text text-to-speech

vocello-design

Historical Vocello 2.0 prototype reference only; not a current production or user-scoped Codex skill.

PowerBeef/Vocello · 27 tokens

add-new-model

Use this skill when the user wants to add or port a new model architecture to MLX-VLM — mapping a Hugging Face modeltype to a new file under mlxvlm/models, writing the ModelConfig, matching layer/weight names, reusing a similar existing model, adding a test class, and validating the port. Covers vision-language…

Blaizzy/mlx-vlm · 83 tokens

benchmarking

Use this skill when the user wants to benchmark an MLX-VLM change and present the numbers in a PR — fork-vs-main A/B comparisons, isolated-module micro-benchmarks, median-of-N timing with warmup, peak-memory reporting, correctness checks, parameter sweeps, and self-contained reproducible bench scripts to paste into a…

Blaizzy/mlx-vlm · 74 tokens

convert-quantize

Use this skill when the user wants to convert a Hugging Face model to MLX or quantize/dequantize one with mlxvlm.convert, including bits and group size, quant modes (affine, mxfp4, nvfp4, mxfp8), RTN vs AWQ, mixed-bit recipes, dtype casts, calibration (text or multimodal), local vs Hub paths, revisions, uploading to…

Blaizzy/mlx-vlm · 99 tokens

cli-inference

Use this skill when the user wants to run or debug MLX-VLM inference from the command line, including uv run mlxvlm.generate, image/audio/video inputs, local model paths, Hugging Face model IDs, deterministic repro commands, and CLI errors around processors, prompts, model loading, or missing weights.

Blaizzy/mlx-vlm · 67 tokens

contributing

Use this skill when the user wants to contribute to MLX-VLM — opening a PR, where model code/config/tests go, backward-compatible config args, running the test suite, code formatting and the pre-commit hooks (black, clang-format), and PR expectations (tests, review, perf evidence). Use it to set up a change so it…

Blaizzy/mlx-vlm · 76 tokens