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 lucastamoios/skills --skill scriptgit clone --depth 1 https://github.com/lucastamoios/skillsWrote 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/lucastamoios/skills/script)<a href="https://agentmods.dev/skills/lucastamoios/skills/script"><img src="https://agentmods.dev/badge/skills/lucastamoios/skills/script/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.
<a href="https://agentmods.dev/skills/lucastamoios/skills/script"><img src="https://agentmods.dev/badge/skills/lucastamoios/skills/script.svg" alt="Reviewed on agentmods" width="80" 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 | $0.00063 | $0.01412 |
| Opus 5.5 | $0.00025 | $0.00565 |
| Sonnet 5.5 | $0.00013 | $0.00282 |
| Haiku 4.5 | $0.00006 | $0.00141 |
Grade A, and why
script 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 4d 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.
How it starts
The opening of the file, as written. The whole thing — 37 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Script
Write what the creator will say and what the camera will show, beat by beat. The script carries the angle's thesis through the reference's mechanism in the creator's voice.
Contract
Follow the marketing contract. Input schema: script.input. Output schema: script.output. Output JSON only.
Generate mode
- Hook (first 3 seconds). Open with what changed for the viewer or what they will get, not with the creator's feat. Show something at the same time. One sentence.
- Promise and so-what early. Within the first 8 seconds the viewer knows what they gain.
2b. Start from the theme and return to it.
creator.themeis the purpose of every video. The hook or the promise puts the viewer inside the theme, and the beat before the CTA brings the fact back to it, in plain words. Videos can differ in subject, example and format, and a beat can drift a little from the theme; what stays the same is that the viewer ends on the theme. A life moment (family, faith, a past job) is a valid way into the theme when the video shows what it says about it. Naming the theme only in the CTA does not count. Do not add a lesson about the theme thatmaterialor the angle does not support. - Structure. Without a reference, use the content expert's default: hook from what changed for the viewer, how it works explained simply, the insight (the "sacada"), CTA.
3b. Carry the mechanism, not the blocks. If
reference_analysisis given, reuse itstransferable_patternand respect itsexecution_requirements. If the creator'sassetscannot meet them, change the visual and record why invisual_alternatives. - Steps need examples when the niche is not visual. A workout list works dry because the body shows each step. In dev or abstract niches, each step is: action, plus an example from the creator's case, plus what it avoided. Prefer fewer steps over vague ones.
4b. Spoken examples are real. Every example in a line comes from
material. Do not invent a demo scenario (a discount rule, a fictional client) and narrate it as if it happened; the viewer cannot tell it apart from the creator's case. If the beat needs an example andmaterialhas none, returnneeds_inputwith a question that asks the creator for one. 4d. Stop instead of writing half a case. If the angle has nooutcome, or a question in itsrisksblocks the video from closing, returnneeds_inputwith that question. Do not write a script that ends before the outcome. 4e. Keep the creator's words. A thesis or opinion of the creator is said with his wording, not softened or extended ("não é" stays "não é", not "não é só"). Any statement about what happened or what he thinks, not only numbers and dates, goes inclaimswith its source. 4c. Connect the facts. Each fact about the creator must connect to the next line and to the thesis. Two loose facts in one sentence (an employer and a team habit) read as one claim the creator did not make. - Every beat has a
visual. The camera always shows something: the creator, a screen, an object, handwritten text, a before and after. Prefer a person on camera with proof on screen. - Simple and spoken. Short sentences, the creator's words from
voice_samplesandbrand, no essay connectors. Read it aloud mentally; if it sounds written, rewrite. - CTA names the audience's pain or desire, then offers the product with where to find it. If
offeris given, build the CTA from one of itscta_linesor from itsenemy,why_noworwhat_to_loseanswers, and pick the one that continues the angle's thesis. Do not invent a freebie, keyword reply, or deadline the creator did not offer. - Claims. Every number, employer, result, or date in a line appears in
claims. Without a source inmaterial,creator.authority, orcreator.brand, use a bracketed placeholder in the line andstatus: "needs_source". Theoffer,angleandcritiqueinputs are not sources: a fact that only appears there staysneeds_source. 8b. Each beat says something new. Never repeat an idea, a date, or a phrase in another beat. Inrevisemode, cutting counts as much as adding: when a new fact comes in, remove what it makes unnecessary. - Length follows the content. Say everything an outsider needs to understand the whole, and nothing that does not serve the idea. Do not cut a needed link to be short; a short script that leaves the viewer without the "what was it?" fails. Back each claim with the concrete thing behind it (the fact, the example, the reason), in plain words. Compute
estimated_secondsat about 2.5 spoken words per second as information only. Iftarget_secondsis given, it is the user's preference, not a limit; if the content clearly holds two ideas, propose a split invisual_alternatives. - Fill
preparationwith what to set up before recording and give twovisual_alternativeswith their trade-offs (e.g. talking head with screen behind; the creator playing two characters).
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.
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.
- 4d ago Changed · +7 lines 791e583395f7
- 11d ago First seen · 30 lines · 63 tokens per session scan A a9d194b18ffb
script is a skill published in the GitHub repository lucastamoios/skills (0 stars, last pushed 5d ago), licensed MIT. It adds 63 tokens to every session and 1,412 once invoked, about $0.0003 per session on Opus 5.5. 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 2026-11-05.
script compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| scriptlucastamoios/skills | Skill | 0 | 5d ago | MIT | A | 63 | Claude Code, Codex |
| imagegenopenai/codex | Skill | 135k | today | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 100k | today | Apache-2.0 | A | 19 | Claude Code |
| browser-automationopenclaw/openclaw | Skill | 395k | today | MIT | A | 36 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 80k | 9d ago | MIT | A | 200 | Claude Code, Codex |
| next-cache-components-adoptionvercel/next.js | Skill | 144k | today | MIT | A | 95 | Claude Code |
| chat-pet-sprite-creationmicrosoft/vscode | Skill | 195k | today | MIT | A | 53 | Claude Code, Codex |