Skip to content
Back to search

human-edge-agent-skills

18 portable Agent Skills for voice-native AI, human judgment, microbets, social tact, writing, life systems, and safe deployment.

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/fairy123456789--human-edge-agent-skills/badge.svg)](/listings/fairy123456789--human-edge-agent-skills)

HTML

<a href="/listings/fairy123456789--human-edge-agent-skills"><img src="/api/listings/fairy123456789--human-edge-agent-skills/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

Install instructions from the README

As written by the repo's author. findsafeskills has not verified these steps.

GitHub CLI 2.90+ supports browsing and installing a specific Agent Skill from a repository:

GitHub CLI:

gh skill preview FAIRY123456789/human-edge-agent-skills vibe-to-spec
gh skill install FAIRY123456789/human-edge-agent-skills vibe-to-spec --agent claude-code --scope user

For another compatible host, use the host supported by your current CLI or copy the individual skill directory into that agent's skills location.

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

.claude-plugin/marketplace.json · Manifest · 84 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. {
  2. "$schema": "https://json.schemastore.org/claude-code-marketplace.json",
  3. "name": "human-edge-skills",
  4. "version": "1.0.0",
  5. "description": "Focused Human Edge Agent Skill plugins for turning messy human context into useful work without flattening judgment or voice.",
  6. "owner": {
  7. "name": "Joy T",
  8. "url": "https://github.com/FAIRY123456789"
  9. },
  10. "plugins": [
  11. {
  12. "name": "voice-to-work",
  13. "displayName": "Voice to Work",
  14. "source": "./plugins/voice-to-work",
  15. "description": "Turn noisy spoken task and product dumps into executable todos or implementation-ready specifications.",
  16. "version": "1.0.0",
  17. "author": {
  18. "name": "Joy T",
  19. "url": "https://github.com/FAIRY123456789"
  20. },
  21. "homepage": "https://github.com/FAIRY123456789/human-edge-agent-skills/tree/main/plugins/voice-to-work",
  22. "repository": "https://github.com/FAIRY123456789/human-edge-agent-skills",
  23. "license": "MIT",
  24. "keywords": ["voice-input", "productivity", "specifications"],
  25. "category": "productivity",
  26. "tags": ["agent-skills", "asr", "task-management"],
  27. "strict": true
  28. },
  29. {
  30. "name": "skill-from-scars",
  31. "displayName": "Skill From Scars",
  32. "source": "./plugins/skill-from-scars",
  33. "description": "Decide whether repeated pain deserves a reusable Agent Skill, then shape it into a testable package.",
  34. "version": "1.0.0",
  35. "author": {
  36. "name": "Joy T",
  37. "url": "https://github.com/FAIRY123456789"
  38. },
  39. "homepage": "https://github.com/FAIRY123456789/human-edge-agent-skills/tree/main/plugins/skill-from-scars",
  40. "repository": "https://github.com/FAIRY123456789/human-edge-agent-skills",
  41. "license": "MIT",
  42. "keywords": ["agent-skills", "retrospectives", "knowledge-capture"],
  43. "category": "development",
  44. "tags": ["skill-authoring", "evaluation", "documentation"],
  45. "strict": true
  46. },
  47. {
  48. "name": "build-in-public-launcher",
  49. "displayName": "Build in Public Launcher",
  50. "source": "./plugins/build-in-public-launcher",
  51. "description": "Turn a finished project into an evidence-first public launch without invented traction or vanity metrics.",
  52. "version": "1.0.0",
  53. "author": {
  54. "name": "Joy T",
  55. "url": "https://github.com/FAIRY123456789"
  56. },
  57. "homepage": "https://github.com/FAIRY123456789/human-edge-agent-skills/tree/main/plugins/build-in-public-launcher",
  58. "repository": "https://github.com/FAIRY123456789/human-edge-agent-skills",
  59. "license": "MIT",
  60. "keywords": ["product-launch", "open-source", "validation"],
  61. "category": "productivity",
  62. "tags": ["build-in-public", "launch", "evidence"],
  63. "strict": true
  64. },
  65. {
  66. "name": "voice-with-temperature",
  67. "displayName": "Voice With Temperature — Preserve Human Writing Voice",
  68. "source": "./plugins/voice-with-temperature",
  69. "description": "Edit written prose for clarity while preserving the writer's lived detail, uncertainty, rhythm, and emotional temperature.",
  70. "version": "1.0.0",
  71. "author": {
  72. "name": "Joy T",
  73. "url": "https://github.com/FAIRY123456789"
  74. },
  75. "homepage": "https://github.com/FAIRY123456789/human-edge-agent-skills/tree/main/plugins/voice-with-temperature",
  76. "repository": "https://github.com/FAIRY123456789/human-edge-agent-skills",
  77. "license": "MIT",
  78. "keywords": ["writing", "editing", "writing-voice"],
  79. "category": "productivity",
  80. "tags": ["prose", "style", "human-voice"],
  81. "strict": true
  82. }
  83. ]
  84. }
README.md · README · 145 lines

View this file on GitHub (read at the repository's current version: the scan did not record which commit it read)

  1. # Human Edge Agent Skills
  2. [![skills.sh installs](https://skills.sh/b/fairy123456789/human-edge-agent-skills)](https://skills.sh/fairy123456789/human-edge-agent-skills)
  3. > **Human judgment where generic AI is weakest: messy voice, social tact, ambiguous decisions, real-world experiments, and safe shipping.**
  4. English · [简体中文](README.zh-CN.md)
  5. AI is increasingly good at generating possibilities. The scarce part is still human: understanding a relationship, repairing noisy speech from context, deciding what matters this week, translating across fields, preserving a real voice, testing ideas against reality, and knowing when a result has enough proof to show the world.
  6. This repository packages those recurring workflows as portable `SKILL.md` folders.
  7. ## Start with four flagship Skills
  8. | Skill | The moment it helps |
  9. |---|---|
  10. | [`vibe-to-spec`](skills/vibe-to-spec/) | You can explain the product aloud, but the coding agent still needs an implementation contract. |
  11. | [`trend-to-microbet`](skills/trend-to-microbet/) | A new AI trend feels urgent, but you need a 48-hour to 7-day test instead of two weeks of FOMO. |
  12. | [`social-signal-and-tact`](skills/social-signal-and-tact/) | The message is factually correct but socially stiff, over-formal, or hard to answer. |
  13. | [`skill-from-scars`](skills/skill-from-scars/) | The same failure happened often enough that it may deserve a tested reusable workflow. |
  14. Each Skill has one job, a trigger-rich `description`, and an independent install path. The rest of the repository is the supporting catalog—not eighteen equally loud claims.
  15. ## The 18 Skills
  16. | Skill | Group | One-line job |
  17. |---|---|---|
  18. | [`social-signal-and-tact`](skills/social-signal-and-tact/) | Social & Communication | Human tact for upward communication |
  19. | [`csdn-technical-writing`](skills/csdn-technical-writing/) | Creator & Writing | Long-form Chinese technical writing without AI micro-paragraphs |
  20. | [`student-money-four-buckets`](skills/student-money-four-buckets/) | Young Life | A four-layer student budget that still leaves room for fun |
  21. | [`voice-dump-to-todo`](skills/voice-dump-to-todo/) | Voice Native | Messy speech in, executable to-do out |
  22. | [`fitness-recovery-coach`](skills/fitness-recovery-coach/) | Young Life | Training plans that respect recovery instead of forcing the calendar |
  23. | [`cross-domain-translator`](skills/cross-domain-translator/) | Judgment & Translation | Translate problem structure across domains, not just terminology |
  24. | [`voice-with-temperature`](skills/voice-with-temperature/) | Creator & Writing | Make writing clearer without washing the human out of it |
  25. | [`aliyun-fullstack-deploy`](skills/aliyun-fullstack-deploy/) | Engineering | Canary-first deployment with proof and rollback |
  26. | [`small-server-fit`](skills/small-server-fit/) | Engineering | Know whether the app fits before buying or deploying |
  27. | [`deployment-proof`](skills/deployment-proof/) | Build in Public | Turn invisible deployment work into credible public evidence |
  28. | [`trend-to-microbet`](skills/trend-to-microbet/) | AI Native Bets | See a trend. Place a small bet. Demand evidence. |
  29. | [`vibe-to-spec`](skills/vibe-to-spec/) | Voice Native | Voice chaos to coding-agent contract |
  30. | [`skill-from-scars`](skills/skill-from-scars/) | AI Native Bets | Turn repeated pain into reusable Agent Skills |
  31. | [`build-in-public-launcher`](skills/build-in-public-launcher/) | AI Native Bets | Launch for evidence, not applause |
  32. | [`personal-boardroom`](skills/personal-boardroom/) | Judgment & Translation | One decision, several explicit lenses, one ranked answer |
  33. | [`anti-slop-personal-brand`](skills/anti-slop-personal-brand/) | AI Native Bets | Make the first screen current, real and earned |
  34. | [`deep-talk-to-podcast`](skills/deep-talk-to-podcast/) | Creator & Writing | Turn a meaningful deep talk into a privacy-aware AI podcast |
  35. | [`one-week-founder-mode`](skills/one-week-founder-mode/) | AI Native Bets | One main bet, two microbets, a kill list |
  36. ## Two voice-native Skills that work together
  37. `voice-dump-to-todo` and `vibe-to-spec` deliberately share a context-aware ASR repair method, but have different destinations:
  38. - **voice-dump-to-todo** → “What should I do?” → actions, decisions, dependencies, due dates, interactive HTML.
  39. - **vibe-to-spec** → “What should the software do?” → scope, user flow, data/state, constraints, acceptance criteria, implementation order.
  40. Both can use recurring named entities from the active conversation/project context to repair obvious ASR errors. They auto-normalize only high-confidence corrections; ambiguous names remain visible instead of being guessed.
  41. ## Plugins
  42. Five of the 18 canonical Skills are also packaged into four focused, self-contained plugins. The `skills/` directories remain the source of truth; packaged copies are generated and hash-checked.
  43. | Plugin | Included Skills |
  44. |---|---|
  45. | [`voice-to-work`](plugins/voice-to-work/) | `voice-dump-to-todo`, `vibe-to-spec` |
  46. | [`skill-from-scars`](plugins/skill-from-scars/) | `skill-from-scars` |
  47. | [`build-in-public-launcher`](plugins/build-in-public-launcher/) | `build-in-public-launcher` |
  48. | [`voice-with-temperature`](plugins/voice-with-temperature/) | `voice-with-temperature` (writing voice, not speech or audio) |
  49. Each directory has a root Agent Plugins 1.0 manifest for portable clients such as Cursor and a Claude Code manifest under `.claude-plugin/`. No plugin adds hooks, agents, commands, MCP servers, or platform-specific behavior.
  50. Current Codex can discover all four from the checked-in repository marketplace:
  51. ```bash
  52. codex plugin marketplace add FAIRY123456789/human-edge-agent-skills --ref main
  53. codex plugin list --marketplace human-edge-skills --available --json
  54. codex plugin add voice-to-work@human-edge-skills
  55. ```
  56. For a local clone, replace the GitHub source in the first command with `.`. The 2026-09-09 Codex runtime record covers discovery, installation, all five packaged Skills, uninstall, and clean reinstall. A marketplace listing is not implied.
  57. Claude Code can add this repository as a marketplace and install one plugin:
  58. ```bash
  59. claude plugin marketplace add FAIRY123456789/human-edge-agent-skills
  60. claude plugin install voice-to-work@human-edge-skills
  61. ```
  62. For a local clone, run `claude plugin validate .`, add the clone with `claude plugin marketplace add . --scope local`, and install with `--scope local`. Cursor's official local-development route is to copy one complete plugin directory into `~/.cursor/plugins/local/<plugin-name>`, reload the window, and confirm its Skills under Customize. These files are packaging and submission preparation; they do not claim a marketplace listing or acceptance.
  63. ## Install one standalone Skill from the monorepo
  64. GitHub CLI 2.90+ supports browsing and installing a specific Agent Skill from a repository:
  65. GitHub CLI:
  66. ```bash
  67. gh skill preview FAIRY123456789/human-edge-agent-skills vibe-to-spec
  68. gh skill install FAIRY123456789/human-edge-agent-skills vibe-to-spec --agent claude-code --scope user
  69. ```
  70. For another compatible host, use the host supported by your current CLI or copy the individual skill directory into that agent's skills location.
  71. ## The deployment arc: fit → ship → prove
  72. The original lightweight `safe-shared-vps-deploy` concept has been deduplicated into the stronger [`aliyun-fullstack-deploy`](skills/aliyun-fullstack-deploy/) workflow. It retains the shared-server protection rules and adds release packaging, canary testing, atomic promotion, deeper verification, and rollback tooling.
  73. - [`small-server-fit`](skills/small-server-fit/) asks whether the app and runtime fit a low-cost server before purchase or deployment.
  74. - [`aliyun-fullstack-deploy`](skills/aliyun-fullstack-deploy/) ships to an authorized Alibaba Cloud ECS or comparable Linux VPS through evidence gates.
  75. - [`deployment-proof`](skills/deployment-proof/) turns the redacted result into an honest public case study and career evidence.
  76. ## Why one repository instead of 18 repositories?
  77. The public entry point stays simple: one GitHub link on a resume or personal website, one star/watch/fork surface, one release history, and one place for issues. Individual Skills still have their own folder, `SKILL.md`, human-facing README, keywords, deep link, and can be installed individually from the same repository.
  78. The repository includes a release strategy in [`docs/MONOREPO_AND_DISCOVERY.md`](docs/MONOREPO_AND_DISCOVERY.md). The recommended rule is: **start as a monorepo; spin out only a Skill that independently earns enough users or contributors to justify its own lifecycle.**
  79. ## Current AI-podcast route
  80. `deep-talk-to-podcast` includes a dated platform selector for source-grounded Audio Overviews, editable AI podcast production, and high-quality generated voices. Because product names and features change quickly, the Skill tells an online agent to verify current official pages before recommending a service.
  81. ## Validation
  82. ```bash
  83. pip install pyyaml
  84. python scripts/validate_skills.py
  85. python scripts/sync_plugin_packages.py --check
  86. python scripts/validate_plugin_packages.py
  87. ```
  88. Seed behavior evals live in [`evals/cases.json`](evals/cases.json).
  89. The NVIDIA SkillEvaluator Tier 1 audit is recorded in [`docs/NVIDIA_TIER1_REPORT.md`](docs/NVIDIA_TIER1_REPORT.md), with the [candidate classification](docs/OFFICIAL_PLUGIN_CANDIDATES.md) and [fix log](docs/TIER1_FIX_LOG.md). Raw JSON is retained under `eval-results/`.
  90. Codex packaging and runtime validation are recorded in the [Codex runtime report](docs/CODEX_RUNTIME_VALIDATION.md). Cross-platform state is summarized in [Cross-Runtime Marketplace Status](docs/CROSS_RUNTIME_MARKETPLACE_STATUS.md). The [Plugin Submission README](docs/PLUGIN_SUBMISSION_README.md) provides one copy-ready source for Claude and Cursor form fields; platform-specific records remain available for [OpenAI](docs/OPENAI_PLUGIN_SUBMISSION_FINAL.md), [Cursor](docs/CURSOR_MARKETPLACE_SUBMISSION_FINAL.md), [Claude](docs/CLAUDE_COMMUNITY_SUBMISSION_FINAL.md), [GitHub Awesome Copilot](docs/GITHUB_AWESOME_COPILOT_SUBMISSION_FINAL.md), [skills.sh](docs/SKILLS_SH_STATUS.md), and the [Gemini CLI feasibility decision](docs/GEMINI_CLI_GALLERY_FEASIBILITY.md). Cursor 3.19.13 discovered all four local plugins and five packaged Skills; its five behavior prompts still require a user visual check. Claude Code validation and runtime remain untested because only Claude Desktop, not the Claude Code CLI, is available on the validation host. No Claude or Cursor Marketplace submission has been made.
  91. ## Privacy and honesty
  92. - no private project names, personal chat logs, contact information, credentials, or company-confidential data;
  93. - no fake metrics, users, testimonials, stars, installs, or adoption;
  94. - no silent low-confidence named-entity correction;
  95. - no fabricated familiarity in social communication;
  96. - no medical diagnosis in fitness/recovery guidance;
  97. - no destructive production action without normal tool permissions and explicit intent.
  98. ## License
  99. MIT.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.