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.
npx skills add jaredlangleycc/nightshift-os --skill voice-from-samplesgit clone --depth 1 https://github.com/jaredlangleycc/nightshift-osWrote 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.
[](https://agentmods.dev/skills/jaredlangleycc/nightshift-os/voice-from-samples)<a href="https://agentmods.dev/skills/jaredlangleycc/nightshift-os/voice-from-samples"><img src="https://agentmods.dev/badge/skills/jaredlangleycc/nightshift-os/voice-from-samples.svg" alt="Measured on agentmods" height="20"></a>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 | $0.00103 | $0.01902 |
| Opus 5 | $0.00051 | $0.00951 |
| Sonnet 5 | $0.00021 | $0.00380 |
| Haiku 4.5 | $0.00010 | $0.00190 |
Grade B, and why
voice-from-samples scanned grade B with 1 finding 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 6d 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.
Strips warnings and disclaimersmediumAnti-refusal
Omitting safety caveats hides risk from the user and is a common jailbreak preamble.
Open the conversation with a one-paragraph explainer. Do not lecture. Example: How it starts
The opening of the file, as written. The whole thing — 121 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Voice From Samples
The voice-capture exercise. Assessment Section 4 asks the user to describe their voice. This skill extracts their voice from real writing. Higher fidelity. Survives self-report bias.
Why this matters
Most people misdescribe their own voice. They say "punchy" when their average sentence is 32 words. They say "warm" when their writing is technical. They say "casual" when they hedge every paragraph.
Real writing samples bypass that gap. Sentence-length distribution does not lie. Signature openers and closers reveal patterns the writer cannot see. Vocabulary preferences expose what the writer actually reaches for, not what they wish they reached for.
Run this skill once after a few weeks of using NightShiftOS when the user has written enough that you can sample real work. Or run it on Day 1 if they already have a corpus they trust (a blog, a newsletter, sent emails).
When this skill triggers
- User says:
/capture-voice, "capture my voice", "learn my style", "voice from writing", "analyze my writing", "what does my voice sound like", "extract my voice" - Optional cron:
voice-recapture-monthly(re-extract every 30 days as the user's writing evolves) - After the user has shipped 5+ pieces of content using NightShiftOS, propose running it
Relationship to voice-rules
voice-rules ENFORCES voice on every output. voice-from-samples POPULATES the voice profile that voice-rules enforces. They pair:
| Skill | When it runs | What it does |
|---|---|---|
customize-from-assessment |
Day 1 onboarding | Populates voice-rules.md from assessment Section 4 (user self-report) |
voice-from-samples (this skill) |
After real writing exists | Updates voice-rules.md with patterns extracted from actual samples (real evidence) |
voice-rules |
Every draft produced by other skills | Enforces the populated voice profile |
Sample-derived voice takes precedence over assessment-derived voice when they conflict. Document both signals in voice-rules.md so the user can see the gap (e.g., "you said 'punchy' but your average sentence is 28 words — sample-derived takes precedence, treat punchy as aspirational").
What ships with it
3 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.
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.
- 6d ago First seen · 121 lines · 103 tokens per session scan B c72ed2c9162f
voice-from-samples is a skill published in the GitHub repository jaredlangleycc/nightshift-os (3 stars, last pushed 1mo ago), licensed MIT. It adds 103 tokens to every session and 1,902 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it B with 1 finding (strips warnings and disclaimers). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
Brand Guidelines Generator
Create and maintain brand style guides for consistent visual identity.
web-vid
Turn a web app or marketing site into a recreated (not screen-recorded) Remotion promo video — pixel-faithful to the site's own design system, choreographed clicks/scrolls/zooms, rendered to MP4. Use when the user asks for a product demo video, a promo/teaser for LinkedIn or social, an "walkthrough video" of their…
pitchdocs
Generate marketing-quality repository documentation from codebase analysis. Scans 10 signal categories, extracts features with file-level evidence, and produces README, CHANGELOG, ROADMAP, and 15+ more docs. Zero runtime dependencies. For AI context file management, see ContextDocs.
demo-video-factory
Generate a custom 26s product-demo video from any SaaS URL — brand-matched scenes, a recreated product-UI "wow" scene, real screenshots, and a soundtrack. Use whenever the user wants a demo, promo, launch, or marketing video for a website or product.
brand
Brand context management for marketing sessions with continuity across conversations.
liking-factor-engineer
Analyze and engineer liking to increase rapport, persuasion, and compliance in marketing, sales, and communication contexts. Use this skill when the user wants to improve how much an audience likes them, their brand, or their message — including writing sales copy, designing onboarding flows, crafting brand voice…