New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Realistic Prompt Generation

by elvisun in elvisun/newsjack

Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands…

MITAuto-check passed

Install Realistic Prompt Generation

skills CLI
$ npx skills add elvisun/newsjack --skill realistic-prompt-generation -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install elvisun/newsjack realistic-prompt-generation --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/elvisun/newsjack.git skills-src && mkdir -p .claude/skills && cp -r skills-src/skills/realistic-prompt-generation .claude/skills/realistic-prompt-generation && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
realistic-prompt-generation
GitHub stars
1.5k
Token cost
~1.6k tokens
SKILL.md length
585 words
Files
1
Skills in repo
30
Repo updated
First seen
Licence
MIT

At a glance

Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands…

  • Works in 2 steps: the closest natural rendering of… → a natural paraphrase that preserves the…
  • SKILL.md covers Hard context boundary, Preserve the canonical intent…, Write realistic prompts and Band rules, plus 2 more sections
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md

What it does

Realistic Prompt Generation is an agent skill from elvisun/newsjack. Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands, and evidence language. Use after architecture design and before contamination or semantic QA.

Its SKILL.md is about 1.6k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

The repository describes itself as: The open-source skills that turn your agent into a full PR team. The licence is MIT.

Example prompts

  • “/realistic-prompt-generation”

Workflow steps

2 steps, taken from the first numbered list in SKILL.md.

  1. the closest natural rendering of observed language;
  2. a natural paraphrase that preserves the same intent.

What it can do on your machine

Read from SKILL.md and the folder at commit b5a8dc8. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md (its code samples are json).

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Realistic Prompt Generation loads about 1.6k tokens when it runs. Until then it costs about 79 tokens; SKILL.md has 585 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~79
When it runs · the whole SKILL.md, loaded when a task matches
~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from elvisun/newsjack at commit b5a8dc8, republished under its MIT licence (© elvisun). 585 words, ~1,641 tokens.

Download SKILL.mdSave it as .claude/skills/realistic-prompt-generation/SKILL.md (or your agent's skills folder).
name
realistic-prompt-generation
description
Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands, and evidence language. Use after architecture design and before contamination or semantic QA.
metadata.category
AI visibility

Realistic Prompt Generation

Write authentic prompts without manufacturing recommendation opportunities.

This skill inherits the ethical floor from skills/ETHICS.md. It enforces anti-hallucination and evidence-bound language. Anti-spray and human-send are not applicable.

Hard context boundary

For unaided generation, accept only:

  • anonymized segment and role labels;
  • approved jobs, constraints, information acts, journey states, and locales;
  • short source-language fragments safe under the permitted-data rules;
  • evidence IDs and grades;
  • prompt_architecture.json;
  • style and turn-form requirements.

Do not accept or inspect:

  • target brands, products, domains, people, slogans, campaign terms, proprietary categories, or flattering claims;
  • current AI answers, rankings, mentions, citations, gaps, or target pages;
  • the contamination register itself.

If those fields appear, stop unaided generation and request a sanitized blind_design_brief.json. When subagents or fresh sessions are available, generate in a fresh context that receives only the blind brief and architecture.

The exception is an explicitly separate B0 aided pass. It may receive only the target aliases needed by approved B0 cells.

Preserve the canonical intent cell

For each architecture cell, hold constant:

  • underlying job;
  • journey state and information act;
  • material constraints;
  • persona/role and locale;
  • proximity band;
  • expected answer kind.

Create two core variants by default:

  1. the closest natural rendering of observed language;
  2. a natural paraphrase that preserves the same intent.

Use additional variants only for a wording-sensitivity pilot or rotating discovery. Do not create a full style × persona × locale grid.

Use variant_role: observed_language only when the candidate is verbatim or lightly normalized from a cited behavioral/query language sample. search_query_expanded, human_written, and llm_expanded candidates are natural_paraphrase or sensitivity; a source ID does not by itself make generated wording observed.

Write realistic prompts

Reflect evidence-supported styles:

  • concise;
  • contextual;
  • imperfect but intelligible;
  • natural follow-up;
  • separately scripted multi-turn only when evidence supports a journey.

Avoid polished persona exposition such as “As a forward-thinking CFO at a 120-person professional-services firm...” Use only context a real person needs to get a useful answer.

Every candidate must record one transformation:

  • verbatim;
  • lightly_normalized;
  • search_query_expanded;
  • human_written;
  • llm_expanded;
  • translated;
  • locale_transcreated.

Do not assign observed frequency to a generated prompt. llm_expanded remains evidence grade D until independently validated or explicitly promoted.

Show full SKILL.md (240 more words)Show less

Band rules

  • Brand (B0) names the supplied target alias and stays target_aided.
  • Shortlist (B1) may ask compare/recommend/buy only when the job evidence supports that act.
  • Category (B2) may name only an accepted evidence-supported category.
  • Problem (B3) supplies the problem/need, not the category.
  • Goal (B4) supplies the outcome/job, not a product or category.
  • Market (B5) needs a fresh evidence ID and a review-by date.

Use the codes in prompt_universe.json and the names in the Markdown generation summary.

Never:

  • append “what tools should I use?” merely to force brands;
  • convert “how do I solve this?” into “which platform should I buy?”;
  • add a product/category not entailed by evidence;
  • use answer-derived wording in core;
  • machine-translate a core locale and call it transcreated;
  • copy prompts from public conversational corpora.

Core non-default locales require native or market-competent review. Until reviewed, set locale_review_status: pending and keep the candidate outside core.

Output

Present a short Markdown generation summary first: cell coverage, variant counts, style mix, grade-D share, locale-review gaps, and any architecture cells that could not be rendered without guessing.

Then write prompt_universe.json:

json
{
  "schema_version": "1.0.0",
  "artifact_id": "universe-<stable-slug>",
  "created_at": "RFC3339",
  "created_by": "declared agent or human",
  "source_manifest_hash": null,
  "warnings": ["hash_not_computed: compute all hashes before freeze"],
  "blind_brief_hash": null,
  "canonical_cells": [
    {
      "canonical_cell_id": "cell-001",
      "cell_spec_id": "spec-001",
      "job_id": "job-001",
      "icp_ids": ["icp-001"],
      "information_act": "diagnose",
      "journey_state": "problem_identification",
      "funnel": null,
      "proximity_band": "B3_problem_need",
      "aided_status": "unaided",
      "campaign_exposed": false,
      "persona_id": "role-001",
      "locale": "en-CA",
      "language": "en",
      "material_constraints": [],
      "expected_answer_kind": "diagnosis_and_options",
      "turn_form": "single_turn",
      "lane_eligibility": ["closed_model", "retrieval"],
      "partition": "core",
      "evidence_grade": "A",
      "reason_source_ids": ["source-001"],
      "candidates": [
        {
          "candidate_id": "prompt-001a",
          "variant_role": "observed_language | natural_paraphrase | sensitivity",
          "text": "Natural user prompt",
          "language": "en",
          "locale": "en-CA",
          "transformation": "lightly_normalized",
          "source_ids": ["source-001"],
          "evidence_grade": "A",
          "locale_review_status": "not_required | pending | approved",
          "generation_provenance": {
            "model": "declared model",
            "prompt_hash": null
          }
        }
      ]
    }
  ]
}

Every canonical cell repeats every flat dimension shown above, copied from its architecture cell. Never shorten the record, move dimensions into a nested object, or add compensating aliases. Do not repair missing jobs or architecture fields. Return unresolved cells to their owning atom.

Handoff

Pass the prompt universe, architecture, safe evidence excerpts, and separately held contamination register to prompt-set-qa. Do not expose baseline visibility.

© elvisun, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

Just SKILL.md in skills/realistic-prompt-generation of elvisun/newsjack.

Open the folder on GitHubat commit b5a8dc8

Compare with similar skills

Realistic Prompt Generation next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Realistic Prompt Generation compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Realistic Prompt Generation this skillelvisun/newsjack1.5k—~1.6kAutomated safety check: PassMIT
Control UIopenclaw/openclaw392k—~1.8kAutomated safety check: PassMIT
Natural Writingflutter/flutter180k—~1.9kAutomated safety check: PassBSD-3-Clause
Control UI E2Eopenclaw/openclaw392k—~2.9kAutomated safety check: PassMIT
Nature Citation FinderYuan1z0825/nature-skills47k—~759Automated safety check: PassApache-2.0
Nature Data AvailabilityYuan1z0825/nature-skills47k—~957Automated safety check: PassApache-2.0

Similar skills

  • Control UI

    openclaw/openclaw

    Operate and troubleshoot the OpenClaw Control UI: navigate connected clients, organize sessions, build session dashboards, and handle direct or Tailscale-hosted Gateways.

    392k GitHub stars~1.8k tokensUpdated today
    Auto-check passed

  • Natural Writing

    flutter/flutter

    Contains well-defined rules for creating natural, accurate, and readable writing.

    180k GitHub stars~1.9k tokensUpdated today
    Writing & ContentAuto-check passed

  • Control UI E2E

    openclaw/openclaw

    A skill your agent uses when designing, testing, fixing, or extending the OpenClaw Control UI GUI, including UI stress-test galleries with feedback inputs, Vitest + Playwright end-to-end checks…

    392k GitHub stars~2.9k tokensUpdated today
    Testing & QAAuto-check passed

  • Nature Citation Finder

    Yuan1z0825/nature-skills

    Finds and verifies Nature-family and CNS literature that supports manuscript claims, maps claims to sources and exports a reference-manager file.

    47k GitHub stars~759 tokensUpdated today
    Research & ScienceAuto-check passed

  • Nature Data Availability

    Yuan1z0825/nature-skills

    Drafts or audits data and code availability statements, dataset access routes, repository plans and FAIR metadata for Nature-style manuscripts.

    47k GitHub stars~957 tokensUpdated today
    Research & ScienceAuto-check passed

  • Bio Variant Calling Structural Variant Calling

    FreedomIntelligence/OpenClaw-Medical-Skills

    Call structural variants (SVs) from short-read sequencing using Manta, Delly, and LUMPY.

    3.1k GitHub stars~1.7k tokensUpdated 2 mo ago
    Research & ScienceAuto-check passed

More from elvisun/newsjack

All 31 skills in this repo

  • Eval Graphics

    elvisun/newsjack

    Turn an eval study's numbers into on-brand, publish-ready figures using the Newsjack chart room (the eval design system), then validate them with Playwright.

    1.5k GitHub stars~2k tokensUpdated today
    Auto-check passed

  • AI Visibility Writing

    elvisun/newsjack

    Audit, question, suggest, or fact-preservingly revise a press release, blog post, contributed article, or expert explainer so AI answer systems can more easily retrieve, understand, quote, and cite…

    1.5k GitHub stars~2.6k tokensUpdated today
    Auto-check passed

  • Coverage Reel

    elvisun/newsjack

    Turn a user-owned coverage CSV and reviewed press-clip captures into an honest, source-linked earned-media dashboard and a motion-designed highlight reel (MP4) that scrolls each real article to the…

    1.5k GitHub stars~3.9k tokensUpdated today
    Auto-check passed

  • Icp Evidence Analysis

    elvisun/newsjack

    Turn a source-bound company and market dossier into testable ideal-customer-profile hypotheses, buying roles, triggers, constraints, disqualifiers, standing, counterevidence, and research gaps.

    1.5k GitHub stars~1.4k tokensUpdated today
    Auto-check passed

  • Build AI Visibility Panel

    elvisun/newsjack

    Research any company, product, or service from a URL plus description and build a comprehensive, evidence-bound AEO/GEO/AI-visibility prompt panel across buyer jobs, information acts, journey…

    1.5k GitHub stars~3.1k tokensUpdated today
    Auto-check passed

  • Reactive Comment

    elvisun/newsjack

    Triage inbound journalist source queries and draft a response only when the user's expertise is a real fit.

    1.5k GitHub stars~5.6k tokensUpdated today
    Auto-check passed

Questions about Realistic Prompt Generation

What does Realistic Prompt Generation do?

Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands…. Realistic Prompt Generation is an agent skill from elvisun/newsjack. Generate natural, controlled prompt variants from a target-blind design brief and prompt architecture while preserving approved jobs, acts, journeys, constraints, roles, locales, proximity bands, and evidence language.

How do I install Realistic Prompt Generation in Claude Code?

Run `npx skills add elvisun/newsjack --skill realistic-prompt-generation -a claude-code`. Or copy the skill folder (skills/realistic-prompt-generation in elvisun/newsjack) into .claude/skills/realistic-prompt-generation in your project. Claude Code loads it when a task matches its description.

How do I install Realistic Prompt Generation in Codex?

Run `npx skills add elvisun/newsjack --skill realistic-prompt-generation -a codex`. Or copy the skill folder (skills/realistic-prompt-generation in elvisun/newsjack) into .agents/skills/realistic-prompt-generation in your project. Codex loads it when a task matches its description.

Can I use Realistic Prompt Generation in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add elvisun/newsjack --skill realistic-prompt-generation -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/realistic-prompt-generation, .gemini/skills/realistic-prompt-generation, .github/skills/realistic-prompt-generation and .opencode/skills/realistic-prompt-generation in your project.

What does Realistic Prompt Generation need to run?

SKILL.md names no scripts, command-line tools or credentials: Realistic Prompt Generation is instructions for the agent only.

Does Realistic Prompt Generation access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Realistic Prompt Generation safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Realistic Prompt Generation use?

Realistic Prompt Generation is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Realistic Prompt Generation use?

About 1.6k tokens (SKILL.md is roughly 6.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Realistic Prompt Generation?

Skills that share tags, products or a category with Realistic Prompt Generation: Control UI (openclaw/openclaw, 392k stars), Natural Writing (flutter/flutter, 180k stars), Control UI E2E (openclaw/openclaw, 392k stars) and Nature Citation Finder (Yuan1z0825/nature-skills, 47k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Realistic Prompt Generation?

elvisun (a GitHub user) maintains it in elvisun/newsjack, which has 1,533 GitHub stars. The repository holds 30 skills in this directory. The repository was last updated on October 7, 2026.

Source: elvisun/newsjack on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.