Skip to content

thatguysam/htma-measure

v0.6.1

A portable collection of agent skills Sam uses and shares.

GitHub

Sam's Skills

Sam's Skills contains reusable workflows for Codex, Claude Code, and other agents that support the Agent Skills format. Install a single skill or the collection bundle. Each skill keeps its instructions in SKILL.md and loads supporting references when needed.

Skills

Andrej Explain

Explain complex ideas, code, research, or agent work in the simplest useful format. Preserve the evidence and limitations while choosing clear text, diagrams, interactive HTML, or narrated walkthroughs. Bundled Node scripts check explicit text constraints and caption timing.

npx skills add thatguysam/skills --skill andrej-explain

Ask: Use $andrej-explain to explain this change, including the mechanism, important boundary case, and what remains untested.

Read the quickstart, examples, research, and verification.

Migrate Custom GPTs

Back up the original details of a user-owned Custom GPT privately, then convert its instructions, knowledge, and integration requirements into a portable skill folder or Git repository. ChatGPT Web delivery uses a supported plugin route. Missing files and Actions remain explicit migration gaps.

npx skills add thatguysam/skills --skill migrate-cgpt

Ask: Use $migrate-cgpt to back up this GPT privately, then convert it into my skills repository.

For ChatGPT Web, private backups, manual folder installation, and revision-pinned installation, see the backup and installation guide.

Read the migration guide and verification limits.

Sales Research

Research sales decisions and draft buyer communication using relevant practitioner sources. Covers discovery, positioning, paid diagnostics, pricing, proposals, AI proof of concept, and sales-call reviews.

npx skills add ThatGuySam/skills --skill sales-research

Ask: Use $sales-research to help me decide the next step in this sales conversation. Include your offer, buyer context, constraints, and the decision you need. No private corpus or companion skill is required.

Read the install and usage guide, examples, and tester checklist.

Sam UX Audit

Audit websites and GitHub repositories through locked UX, accessibility, HIG, motion, and performance review skills. Findings include evidence, scope limits, and actionable acceptance checks.

npx skills add thatguysam/skills --skill sam-ux-audit

Ask: Use $sam-ux-audit to audit this URL or GitHub repository. The agent restores the skill's locked dependencies on first use. Request a dependency update explicitly when you want the latest upstream instructions.

Read the guide and smoke evals.

Zach Prompting

Improve prompts, skills, system instructions, agent definitions, tool descriptions, CLAUDE.md, and AGENTS.md files without weakening their intent or safeguards.

npx skills add thatguysam/skills --skill zach-prompting

Invoke a standalone installation as /zach-prompting in Claude Code or $zach-prompting in Codex.

Example prompts:

Use $zach-prompting to tighten this AGENTS.md while preserving its safety,
permission, validation, and repository-boundary rules.

Use $zach-prompting to migrate this working prompt to GPT-6 Astra one measured
change at a time.

Read the Zach Prompting guide.

HTMA Measure

Estimate uncertain costs, budgets, rates, risks, ROI, and market size with ranges, sources, and recommendations.

npx skills add thatguysam/skills --skill htma-measure

Situations it helps with:

  • you need a planning budget before vendor quotes exist;
  • a stakeholder wants one number even though the evidence supports a range;
  • unfamiliar work or hidden subcomponents keep making estimates too optimistic;
  • "cost" could mean public price, budget allowance, likely quote, or paid amount;
  • only a few observations are available, but ignoring them would waste evidence; or
  • research keeps expanding without a rule for when to decide.

Example prompts:

Use $htma-measure to create a planning allowance for this project before quotes arrive.
Use $htma-measure to turn this point estimate into a calibrated decision range.
Use $htma-measure to estimate this unfamiliar project by decomposing the hidden work.
Use $htma-measure to update our risk estimate from these few observations.
Use $htma-measure to tell us whether more research could still change the decision.

Read the HTMA Measure README, guide, and source-backed problems it solves.

Review or rerun the HTMA Measure evaluation suite.

Documentation

Other installation paths

Clone the repository and copy the required directory from skills/ into your agent's skills directory:

git clone https://github.com/thatguysam/skills.git

The repository also ships version 0.6.1 of a marketplace bundle for Codex and Claude Code. The bundle loads every published skill; use the Skills CLI when you want only one. Its stable install ID remains htma-measure, while Claude Code uses the collection-level sam command namespace.

Codex

codex plugin marketplace add thatguysam/skills
codex plugin add htma-measure@thatguysam-skills

Claude Code

/plugin marketplace add thatguysam/skills
/plugin install htma-measure@thatguysam-skills

Claude Code namespaces bundled skills with sam, for example /sam:zach-prompting and /sam:htma-measure. Standalone installations keep their unbundled names. Codex skill selectors remain $zach-prompting and $htma-measure.

Repository structure

skills/
  andrej-explain/
    SKILL.md
    agents/openai.yaml
    references/
    scripts/
  migrate-cgpt/
    SKILL.md
    agents/openai.yaml
    references/
    assets/
    scripts/
  sales-research/
    SKILL.md
    README.md
    agents/openai.yaml
    references/
    evals/
  sam-ux-audit/
    SKILL.md
    package.json
    package-lock.json
    skills-lock.json
    scripts/
    references/
    evals/
  htma-measure/
    README.md
    SKILL.md
    agents/openai.yaml
    assets/
    evals/
    references/
  zach-prompting/
    SKILL.md
    agents/openai.yaml
    references/vendor-guidance.md

Each skill keeps discovery metadata and the core workflow in SKILL.md, with conditional detail in supporting resources.

Acknowledgments

  • Andrej Explain is inspired by Andrej Karpathy’s writing about understanding model outputs and the sources in its research memo. It is not affiliated with or endorsed by him.

  • Sales Research uses original workflow instructions informed by the practitioners linked in its source guide. Their work remains with its publishers; inclusion does not imply endorsement.

  • Sam UX Audit composes the referenced skills from Addy Osmani, Emil Kowalski, and the community HIG reviewer by dickwu using Luisa Lima’s skills-lock. The HIG reviewer is not an Apple-authored package; each dependency retains its own provenance and terms.

  • Zach Prompting's name and original spark were inspired by Zach Miles. Its concrete prompting methods synthesize the cited OpenAI and Anthropic guidance; it is not affiliated with or endorsed by them.

  • HTMA Measure is inspired by measurement and uncertainty-management ideas popularized by Douglas W. Hubbard's How to Measure Anything. It is not affiliated with or endorsed by the author or publisher.

Contributing

Use trunk-based development for ordinary authorized changes: validate and publish directly to main, with branches or pull requests for explicit requests or required repository rules. Changes should keep each skill specific, portable, verifiable, and minimal. See CONTRIBUTING.md.

License

MIT. See LICENSE.