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

New Feature

by joselado in joselado/pyqula

The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it.

GPL-3.0Auto-check passed

Install New Feature

skills CLI
$ npx skills add joselado/pyqula --skill new-feature -a claude-code

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

GitHub CLI
$ gh skill install joselado/pyqula new-feature --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/joselado/pyqula.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/new-feature .claude/skills/new-feature && 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
new-feature
GitHub stars
145
Token cost
~1.4k tokens
SKILL.md length
768 words
Files
1
Skills in repo
6
Repo updated
First seen
Licence
GPL-3.0

At a glance

The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it.

  • Works in 5 steps: Implementation in a *tk/ subpackage, or… → A one-line delegator on Hamiltonian or… → Follow the mutate-and-return convention.… → …
  • Adding a new method
  • SKILL.md covers Before writing code, Where the code goes, The test and The five documentation surfaces, plus 2 more sections
  • Calls python

What it does

New Feature is an agent skill from joselado/pyqula. The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it. Use when adding a new method, formalism, observable, or Hamiltonian term, or when asked whether a feature is finished.

Its SKILL.md is about 1.4k 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: Python library to compute properties of quantum tight binding models, including topological, electronic and magnetic properties and including the effect of many-body interactions. The licence is GPL-3.0.

When your agent uses it

  • Adding a new method
  • Hamiltonian term
  • Asked whether a feature is finished

Example prompts

  • “/new-feature”

Requirements

  • Python 3

Workflow steps

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

  1. Implementation in a *tk/ subpackage, or in the top-level module that
  2. A one-line delegator on Hamiltonian or Geometry if it is meant to be
  3. Follow the mutate-and-return convention. Methods that add terms modify
  4. Guard arguments with a real message. ValueError for a bad value or a
  5. Parallelism: prefer numba @jit(parallel=True)/prange over

What it can do on your machine

Read from SKILL.md and the folder at commit a61709a. 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

    Shell commands in SKILL.md call:

    • python

    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

New Feature loads about 1.4k tokens when it runs. Until then it costs about 80 tokens; SKILL.md has 768 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~80
When it runs · the whole SKILL.md, loaded when a task matches
~1.4k

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 joselado/pyqula at commit a61709a, republished under its GPL-3.0 licence (© joselado). 768 words, ~1,408 tokens.

Download SKILL.mdSave it as .claude/skills/new-feature/SKILL.md (or your agent's skills folder).
name
new-feature
description
The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it. Use when adding a new method, formalism, observable, or Hamiltonian term, or when asked whether a feature is finished.

Adding a feature to pyqula

A feature here is not done when the code runs. It is done when a stranger can find it, a test pins what it promises, and nothing in the docs contradicts it. This is the checklist for that. CLAUDE.md points here and no longer states these rules itself, so this file is their only home.

Before writing code

  • Is it a new formalism or an extension of one already here? For anything genuinely new -- a method the codebase does not already implement -- check arXiv for a paper covering it and use that reference to guide the implementation, rather than general knowledge alone. Fetch the e-print TeX source, not the HTML: the HTML mangles prefactors and drops index conventions. Re-derive every prefactor yourself.
  • Is there a tested open-source implementation under a compatible license? Benchmark against it, or mirror its structure, rather than writing the algorithm from scratch. (wanniertk/wannierpy/ is the precedent: a bundled pure-Python port rather than a reimplementation.)
  • Grep examples/ first. There may already be a script doing most of it.
  • Check future_development/README.md -- if a roadmap covers the area, it probably records a measurement or a dead end you would otherwise re-derive.

Where the code goes

  1. Implementation in a *tk/ subpackage, or in the top-level module that composes one (topology.py over topologytk/, scf.py over scftk/). Non-trivial functionality does not live in hamiltonians.py.
  2. A one-line delegator on Hamiltonian or Geometry if it is meant to be called as h.get_something(...). The class is deliberately thin; the delegator does nothing but call into the module.
  3. Follow the mutate-and-return convention. Methods that add terms modify in place and return self; callers .copy() before mutating. Do not quietly make a new method purely functional if its siblings are not.
  4. Guard arguments with a real message. ValueError for a bad value or a Hamiltonian in the wrong Hilbert space, NotImplementedError for a combination not built yet, TypeError for a wrong type -- each naming the offending value. A string-selected option (mode=, solver=, channel=) must list the accepted values in the error. Never a bare raise outside the jump-to-except idiom.
  5. Parallelism: prefer numba @jit(parallel=True)/prange over parallel.pcall's process pool. The pool measured slower than serial on the KPM moment loop. Reach for pcall only when the work is not numba-jittable.
Show full SKILL.md (387 more words)Show less

The test

tests/<topic>/test_*.py, asserting a physical or numerical invariant, not a recorded number. The suite's standard shapes:

  • a result must not depend on something it physically cannot depend on (the random seed of an SCF initial guess, the choice of unit cell);
  • two independent code paths computing the same quantity agree to tolerance;
  • a symmetry or sum rule holds (a Goldstone mode for a spin response, a quantized invariant, a conserved current).

Traps that make a correct calculation look broken, or a broken one look correct:

  • sum(bands) == 0 and friends are vacuous for a symmetric spectrum -- the assertion passes whatever the code does. Assert something that can fail.
  • Check nk-convergence before asserting a nonzero value. A hung or wrong mean-field result is usually the k-mesh, not the mixing.
  • Test a generic direction, never one axis. full_dm is the transpose of rho, and contracting it flips the sign of sy, valley and current operators -- an x-only test cannot see it.
  • SCF failure is a None return, not an exception, and maxite defaults to None so a non-converging loop never returns.

The five documentation surfaces

All of these, for a user-facing feature:

  1. documentation/user_guide.md -- a prose section with the physics and motivation plus a runnable snippet, in the existing style.
  2. documentation/user_guide.md, # Main functions and methods -- an entry, for anything with a method on Hamiltonian or Geometry.
  3. README.md, the # FUNCTIONALITIES # list -- a bullet where relevant.
  4. examples/<dimensionality>/<name>/main.py -- a runnable script. These double as usage documentation and are where the next person will grep.
  5. jupyter-notebooks/functionalities/ -- a notebook if the FUNCTIONALITIES bullet should link to one. The README's tutorial section states how many bullets currently do; if you add a notebook, that count moves with it.

Then run python -m pytest tests/documentation -- it statically checks that every method the guide names actually exists -- and rebuild the PDF. The refresh-docs skill does that sweep.

If you deliberately leave something unbuilt

Write it up in future_development/, with what was measured and what the next decision point is, and add it to that directory's README.md index. The point is that picking the work up again does not mean re-deriving a conclusion someone already reached.

Never

Put cluster details -- hostnames, scratch paths, partitions, job IDs, queue measurements -- into anything tracked. Performance conclusions belong in the roadmaps; the machine that produced them does not.

© joselado, GPL-3.0. 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 .claude/skills/new-feature of joselado/pyqula.

Open the folder on GitHubat commit a61709a

Compare with similar skills

New Feature 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.

New Feature compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
New Feature this skilljoselado/pyqula145—~1.4kAutomated safety check: PassGPL-3.0
Verification Before Completionforyourhealth111-pixel/Vibe-Skills3.6k—~1.1kAutomated safety check: PassApache-2.0
Verification Before Completionfarm-fe/farm5.6k46 repos~1kAutomated safety check: PassMIT
Hugging Face Evaluationsickn33/agentic-awesome-skills47k2 repos~418Automated safety check: PassMIT
Hugging Face Datasetssickn33/agentic-awesome-skills47k2 repos~1.1kAutomated safety check: PassMIT
Hugging Face Paperssickn33/agentic-awesome-skills47k1 repos~2.5kAutomated safety check: PassApache-2.0

Similar skills

  • Verification Before Completion

    foryourhealth111-pixel/Vibe-Skills

    Completion-evidence route used before claiming work is complete, fixed, passing, committed, or PR-ready.

    3.6k GitHub stars~1.1k tokensUpdated 1 mo ago
    Agent WorkflowsAuto-check passed

  • Verification Before Completion

    farm-fe/farm

    A skill your agent uses when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any…

    5.6k GitHub starsUsed in 46 repos~1k tokens
    Agent WorkflowsAuto-check passed

  • Hugging Face Evaluation

    sickn33/agentic-awesome-skills

    Add and manage evaluation results in Hugging Face model cards.

    47k GitHub starsUsed in 2 repos~418 tokens
    AI & LLM EngineeringAuto-check passed

  • Hugging Face Datasets

    sickn33/agentic-awesome-skills

    Create and manage datasets on Hugging Face Hub. An agent skill from sickn33/agentic-awesome-skills.

    47k GitHub starsUsed in 2 repos~1.1k tokens
    AI & LLM EngineeringAuto-check passed

  • Hugging Face Papers

    sickn33/agentic-awesome-skills

    Look up and read Hugging Face paper pages in markdown, and use the papers API for structured metadata such as authors, linked models/datasets/spaces, Github repo and project page.

    47k GitHub starsUsed in 1 repo~2.5k tokens
    AI & LLM EngineeringAuto-check passed

  • Verification Before Completion

    jnMetaCode/superpowers-zh

    Chinese-language rule that bars an agent from claiming work is done, fixed or passing until it has run a verification command and read the output.

    8.3k GitHub stars~443 tokensUpdated 2 days ago
    Agent WorkflowsAuto-check passed

More from joselado/pyqula

  • Refresh Docs

    joselado/pyqula

    Refresh pyqula's documentation after a change - recount the test suite, propagate every number that moved, re-run the static user-guide checks, and rebuild documentation/userguide.pdf.

    145 GitHub stars~816 tokensUpdated 3 days ago
    Auto-check passed

  • Error Conventions

    joselado/pyqula

    How pyqula raises errors -- which exception type for which failure, the registries behind string-selected options (mode=, solver=, channel=, operator names), and the shared Hilbert-space guards in…

    145 GitHub stars~1.1k tokensUpdated 3 days ago
    Auto-check passed

  • GPU Backend

    joselado/pyqula

    pyqula's CPU/GPU switch (src/pyqula/gpu.py), how a routine is routed onto the device, per-call precision, and the tiered porting plan in documentation/gpuportingplan.md.

    145 GitHub stars~679 tokensUpdated 3 days ago
    Auto-check passed

  • User Guide Voice

    joselado/pyqula

    The maintainer's writing voice for documentation/userguide.md -- the three registers (chapter prose, section intros, catalogue bullets), the spelling decisions, what not to write, and which chapters…

    145 GitHub stars~985 tokensUpdated 3 days ago
    Auto-check passed

  • Wannierization

    joselado/pyqula

    pyqula's Wannierization (wanniertk/), what h.getwannierhamiltonian() returns, the disentanglement window keywords and which combinations raise NotImplementedError, and the bundled pure-Python…

    145 GitHub stars~1.2k tokensUpdated 3 days ago
    Auto-check passed

Questions about New Feature

What does New Feature do?

The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it. New Feature is an agent skill from joselado/pyqula. The completeness checklist for adding a user-facing feature to pyqula - where the implementation goes, what kind of test it needs, and the five documentation surfaces that must move with it.

When should I use New Feature?

New Feature fits situations like: adding a new method; hamiltonian term; asked whether a feature is finished.

How do I install New Feature in Claude Code?

Run `npx skills add joselado/pyqula --skill new-feature -a claude-code`. Or copy the skill folder (.claude/skills/new-feature in joselado/pyqula) into .claude/skills/new-feature in your project. Claude Code loads it when a task matches its description.

How do I install New Feature in Codex?

Run `npx skills add joselado/pyqula --skill new-feature -a codex`. Or copy the skill folder (.claude/skills/new-feature in joselado/pyqula) into .agents/skills/new-feature in your project. Codex loads it when a task matches its description.

Can I use New Feature 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 joselado/pyqula --skill new-feature -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/new-feature, .gemini/skills/new-feature, .github/skills/new-feature and .opencode/skills/new-feature in your project.

What does New Feature need to run?

Going by SKILL.md and its folder, New Feature needs the command-line tools its instructions call (python). Our summary lists: Python 3.

Does New Feature 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 New Feature 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 New Feature use?

New Feature is published under the GPL-3.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does New Feature use?

About 1.4k tokens (SKILL.md is roughly 5.6k 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 New Feature?

Skills that share tags, products or a category with New Feature: Verification Before Completion (foryourhealth111-pixel/Vibe-Skills, 3.6k stars), Verification Before Completion (farm-fe/farm, 5.6k stars), Hugging Face Evaluation (sickn33/agentic-awesome-skills, 47k stars) and Hugging Face Datasets (sickn33/agentic-awesome-skills, 47k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains New Feature?

joselado (a GitHub user) maintains it in joselado/pyqula, which has 145 GitHub stars. The repository holds 6 skills in this directory. The repository was last updated on October 7, 2026.

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