Agent skill

Contributing

by Dpro-at in Dpro-at/Tel-Agent

A skill your agent uses when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking…

AGPL-3.0Auto-check: notesDevelopment

Install Contributing

skills CLI
$ npx skills add Dpro-at/Tel-Agent --skill contributing -a claude-code

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

GitHub CLI
$ gh skill install Dpro-at/Tel-Agent contributing --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/Dpro-at/Tel-Agent.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.claude/skills/contributing .claude/skills/contributing && 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
contributing
GitHub stars
1.1k
Used in
1 other repo
Token cost
~3.3k tokens
SKILL.md length
1,748 words
Files
1
Skills in repo
2
Repo updated
First seen
Licence
AGPL-3.0

At a glance

A skill your agent uses when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking…

  • Works in 7 steps: Find something to work on → Claim it → Set up, once → …
  • Contributing to Tel-Agent - picking an issue to work on
  • SKILL.md covers Before you touch anything, Translation — the first…, Step 1 — Find something to… and Step 2 — Claim it, plus 7 more sections
  • Calls gh, git and npm; reaches cli.github.com

What it does

Contributing is an agent skill from Dpro-at/Tel-Agent. Use when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking "what can I work on" / "how do I start" / "is my change ready to submit". Covers the full path from a fresh clone to a merged pull request.

Its SKILL.md is about 3.3k tokens, which your agent loads only when the skill is triggered. It is a single SKILL.md file with no bundled scripts.

It sits in Development, covering Pull requests. The repository describes itself as: AI phone assistant | open-source. The licence is AGPL-3.0.

When your agent uses it

  • Contributing to Tel-Agent - picking an issue to work on
  • Setting up the repository for the first time
  • Finishing a task
  • Opening a pull request

Example prompts

  • “what can I work on”
  • “how do I start”
  • “is my change ready to submit”
  • “/contributing”

Requirements

  • Python 3
  • Node.js

Workflow steps

7 steps, taken from the step headings in SKILL.md.

  1. Find something to work on
  2. Claim it
  3. Set up, once
  4. Do the work
  5. Verify before you open anything
  6. Open the pull request
  7. Review

What it can do on your machine

Read from SKILL.md and the folder at commit 8a5193b. 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:

    • gh
    • git
    • npm
    • node
    • npx
    • ruff
    • python
    • mypy
    • pytest

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

  • Network

    Hosts in commands or code, which the agent is likely to contact:

    • cli.github.com

    Also links to:

    • github.com

    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

Contributing loads about 3.3k tokens when it runs. Until then it costs about 82 tokens; SKILL.md has 1,748 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~82
When it runs · the whole SKILL.md, loaded when a task matches
~3.3k

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: notes

The automated check noted patterns worth knowing about, such as sudo or a known installer.

  • NoteMentions a .env fileSKILL.md:26
    needed to run something, it goes in `.env`, which is ignored.

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 Dpro-at/Tel-Agent at commit 8a5193b, republished under its AGPL-3.0 licence (© Dpro-at). 1,748 words, ~3,311 tokens.

Download SKILL.mdSave it as .claude/skills/contributing/SKILL.md (or your agent's skills folder).
name
contributing
description
Use when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking "what can I work on" / "how do I start" / "is my change ready to submit". Covers the full path from a fresh clone to a merged pull request.

Contributing to Tel-Agent

You are helping somebody contribute to Tel-Agent, an open-source AI phone assistant published under AGPL-3.0 by Dpro GmbH.

Read CLAUDE.md at the repository root before anything else. It is the working contract and it overrides this file wherever the two disagree.

Before you touch anything

Three rules that are not negotiable, and breaking any of them wastes the maintainer's time and the contributor's:

  1. Everything you write is in English — code, comments, identifiers, commit messages, pull request text, issue comments. This is Rule 0 in CLAUDE.md. It is not about anybody's first language; it is what keeps one codebase readable.
  2. Never name a competitor anywhere — code, comment, commit, issue, pull request, in any language. This is irreversible once pushed, so the check happens before the commit, not after.
  3. Never commit a secret — no key, token, password, or real phone number. If one is needed to run something, it goes in .env, which is ignored.

Translation — the first contribution, and the one that never runs out

Do not send somebody new to a code issue. Send them here.

If this is their first contribution, walk them through docs/ONBOARDING.md rather than this section. It is the same work written for somebody who has never opened a pull request: one file, about thirty strings, no setup at all. This section is the reference; that page is the path.

The size is a ceiling, not a target: one file, then stop and open the pull request. The point of a first contribution is not the translation in it — it is running the fork-to-merge loop once on something small enough that a mistake costs nothing. A contributor who translates thirty strings and opens a pull request has learned more than one who translates three hundred and never opens it.

Thirty languages sit in locales/ as directories holding nothing but .gitkeep — that is the pick-list, and the script prints it. Any language not among them is added in the same pull request as its first translated file.

The whole task is one loop, and it is the same loop every time:

run the script  →  pick what is missing  →  translate it  →  open a pull request
       ↑                                                              │
       └──────────────────────────────────────────────────────────────┘
bash
node scripts/check-locales.mjs               # every language, and where each stands
node scripts/check-locales.mjs --locale fr   # one language: what is left, smallest first

The script measures everything against English, which is the source language. It prints the remaining files smallest first, and the smallest are 16 to 26 strings — about twenty minutes.

bash
mkdir -p locales/fr
cp locales/en/code.json locales/fr/
# translate the values, leave the keys alone
node scripts/check-locales.mjs --locale fr   # confirm the number moved

Then the normal flow from step 2: claim, branch, pull request. One file per pull request. A translator who has finished one runs the script again, takes the next file — or a different language entirely — and repeats. Nothing needs to be assigned again after the first time.

This is why it never runs out: every language is 2,972 strings across 33 files, and any language anybody speaks is welcome. There is more of this work than there will ever be people to do it.

Four things that get a translation pull request sent back, all easy to avoid:

  • Copying all 33 files and translating one. An untranslated copy of English looks finished and ships English to somebody who asked for their own language. One file.
  • Translating the keys. Only the values on the right-hand side change.
  • Translating {placeholders}. {count} and {name} are filled at runtime, and a translated placeholder breaks only on that language — so nobody catches it.
  • Machine translation nobody read. Translate the meaning. If it reads oddly to a native speaker it is wrong, whatever the dictionary says. Only take a language you actually speak.

locales/README.md has the rest: the two tiers, why a partial language breaks nothing, and when a language gets registered.

After the first one is merged, anything on the board is open to them. The translation is not a test to be passed — it is the smallest real thing in the project, and it is where the workflow gets learned on something that does not matter if it takes two attempts.


Step 1 — Find something to work on

Work comes from the public board and nowhere else. Do not invent a task, and do not "just fix something you noticed while you were in there" — that goes in a new issue.

bash
gh issue list --repo Dpro-at/Tel-Agent --label "good first issue" --state open
gh issue list --repo Dpro-at/Tel-Agent --label "help wanted" --state open

The board is at https://github.com/orgs/Dpro-at/projects/6

Only take from the Ready column. Backlog means blocked on something else; the issue looks available and is not. Every issue carries a level: label:

LabelWhat it means
level: first-issueSmall, self-contained, finishable in one sitting. Start here.
level: easyStraightforward, little context needed.
level: mediumNeeds some understanding of the codebase.
level: hardSubstantial, touches several parts.
level: maintainerDecides architecture. Do not take this one.

If the person you are helping has not contributed here before, steer them to level: first-issue. There are real ones, not a token entry.

Read the whole issue before starting. Every issue is written with the same fields: Why (the reason it exists), Do (what to build), Done when (the acceptance test), Verify (the exact commands a reviewer will run), and sometimes Needs (issues that must be closed first). If Needs lists something still open, this task is not ready — go back to step 1.

Step 2 — Claim it

bash
gh issue comment <number> --repo Dpro-at/Tel-Agent --body "I would like to take this."

Wait to be assigned before writing code. One person per issue, and the assignment is what stops two people building the same thing. If nobody responds within a few days, say so on the issue rather than starting anyway.

Step 3 — Set up, once

If the person you are helping has never worked on this repository, walk them through this in order and confirm each command actually worked before moving on. Do not paste all of it at once and hope. Budget about half an hour the first time, and roughly nothing every time after.

3.1 — What has to be installed
bash
git --version      # any recent version
node --version     # v20 or newer. Next.js 16 and React 19 will not run on v18
npm --version
gh --version       # GitHub CLI - https://cli.github.com
python --version   # 3.12+, only needed for backend issues

gh is not optional here. Claiming an issue, opening the pull request and reading the board all go through it. If it is missing, install it before anything else.

Python is only needed if the issue touches api/ or agent/. A documentation or frontend issue needs Node and nothing more.

3.2 — Log in to GitHub
bash
gh auth status

If that says you are not logged in:

bash
gh auth login

Choose GitHub.com → HTTPS → Yes to authenticate git → Login with a web browser, then paste the one-time code it prints.

This step is skipped more often than any other, and it fails late rather than early: everything looks fine until the first gh command, which then prints "To get started with GitHub CLI, please run: gh auth login" instead of doing the thing. If any gh command produces that message, this is why.

Reading the project board needs one extra scope, which the default login does not grant:

bash
gh auth refresh -s read:project
Show full SKILL.md (659 more words)Show less
3.3 — Fork and clone

Everyone contributes through a fork. Nobody gets write access to the repository, and that is the same rule for everyone.

bash
gh repo fork Dpro-at/Tel-Agent --clone
cd Tel-Agent
git remote -v          # expect: origin = your fork, upstream = Dpro-at

gh repo fork --clone sets upstream for you. If git remote -v does not show it:

bash
git remote add upstream https://github.com/Dpro-at/Tel-Agent.git
3.4 — Check the email on your commits
bash
git config user.email

If that address is not on your GitHub account — or your @users.noreply.github.com one — your commits land unlinked: the work gets merged and your name is on nothing.

This is the single most common way somebody does real work here and does not appear in the contributors list, and it is awkward to correct afterwards. Fix it now:

bash
git config user.email "you@example.com"
3.5 — Run what exists

The frontend runs today:

bash
npm --prefix web install
npm --prefix web run dev

Open http://localhost:3000. It redirects to a locale — /en, /de or /ar — and you should land on the sign-in screen. If you see that, the setup worked. Every screen has a state switcher pinned to the top in development; it is stripped from a production build.

The screens read from static fixture modules, so they render without a server. Nothing is wired to anything yet.

The backend does not run, because it does not exist. agent/ and api/ hold six .gitkeep files and two READMEs, and building them is what most of the open issues are. If an issue asks you to run something that is not there, re-read it — the task is almost certainly to create it.

3.6 — Confirm the tooling before you touch code
bash
npx --prefix web tsc --noEmit    # expect: no output
npm --prefix web run lint

Run these before your first edit, not after. If something is already red on a clean clone, that is worth reporting as its own issue — and it is not yours to fix inside an unrelated pull request.

Step 4 — Do the work

bash
git checkout -b <type>/<short-name>

Types: feat/, fix/, docs/, chore/, refactor/, test/. Example: feat/backend-skeleton.

Stay inside the issue. A pull request that does two things gets reviewed as slowly as its harder half. Something else that needs fixing goes in a new issue.

Commits follow Conventional Commits:

feat(api): add health endpoint
docs(contributing): add a what-to-work-on section

Match the code around you — its naming, its comment density, its idiom. A change that reads like it was always there is the goal.

Step 5 — Verify before you open anything

Run the issue's own Verify commands first, then the repository gate for whichever half you touched:

bash
# Python
ruff check . && ruff format --check . && mypy . && pytest

# web/
npm --prefix web run lint
npx --prefix web tsc --noEmit
npm --prefix web run build

Nothing you touched may go red.

Then check by hand:

  • No secret, key, token, password or real phone number in the diff.
  • No competitor named anywhere, in any language.
  • Everything in English.
  • If it is visible, screenshots in light and dark, and one in ar if the layout is direction-sensitive.

Do not report the task finished until these actually pass. If something fails and you cannot fix it, say so plainly in the pull request — a known failure stated honestly is useful; a silent one wastes a review cycle.

Step 6 — Open the pull request

bash
git push -u origin <branch>
gh pr create --repo Dpro-at/Tel-Agent --title "<same as the commit>" --body "..."

The body says: what you did, what you deliberately did not do, and how you tested it. Link the issue with Closes #<number>. Paste the output of the Verify commands.

Open it as a draft while it is not ready. Ready for review means you believe it is done.

The CLA. The first pull request cannot be merged until its author has signed the Contributor License Agreement. Expect to be asked.

Step 7 — Review

Expect changes to be requested. It is not a verdict on the person.

Push follow-up commits to the same branch. Do not force-push during a review — the reviewer loses their place in the diff.

It lands as a squash-merge, done by a maintainer. The branch is deleted automatically.

If the work has to stop

Half the people who take an issue never finish, and that is normal. Saying so is a contribution: it returns the task to the pool the same day instead of fourteen days later, which is when an inactive assignment is cleared automatically.

bash
gh issue comment <number> --repo Dpro-at/Tel-Agent --body "I cannot finish this - unassigning so somebody else can take it."

What "done" means

The code works, the tests cover it, the documentation says so, and .env.example lists any variable that was introduced. All four, not three.

© Dpro-at, AGPL-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/contributing of Dpro-at/Tel-Agent.

Open the folder on GitHubat commit 8a5193b

Used in 1 other repository

We found 1 copy of this SKILL.md (exact, near-identical or edited) in other folders, from 1 other GitHub owner. This page covers the copy in Dpro-at/Tel-Agent, which our catalogue first saw on October 7, 2026.

Compare with similar skills

Contributing 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.

Contributing compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Contributing this skillDpro-at/Tel-Agent1.1k1 repos~3.3kAutomated safety check: NotesAGPL-3.0
Finishing a Development Branchobra/superpowers296k5 repos~1.9kAutomated safety check: PassMIT
PR Babysitteropeninterpreter/openinterpreter69k3 repos~4.2kAutomated safety check: PassApache-2.0
Check PRonyx-dot-app/onyx32k2 repos~2.3kAutomated safety check: PassMIT
Understand Diff AnalysisEgonex-AI/Understand-Anything86k1 repos~1.4kAutomated safety check: PassMIT
PR Design DocOpenHands/OpenHands90k—~2.4kAutomated safety check: PassMIT

Similar skills

  • Walks the last step of a branch: confirm tests pass, detect the git environment, ask how to integrate, carry out your choice and clean up the worktree.

    296k GitHub starsUsed in 5 repos~1.9k tokens
    DevelopmentAuto-check passed
  • PR Babysitter

    openinterpreter/openinterpreter

    Watches an open GitHub pull request until it merges, handling review comments, diagnosing CI failures and retrying flaky checks along the way.

    69k GitHub starsUsed in 3 repos~4.2k tokens
    DevelopmentAuto-check passed
  • Check PR

    onyx-dot-app/onyx

    Checks a GitHub, GitLab, or Perforce (p4) pull request (or merge request, or shelved changelist) for unresolved review comments, failing status checks, and incomplete PR descriptions.

    32k GitHub starsUsed in 2 repos~2.3k tokens
    DevelopmentAuto-check passed
  • Understand Diff Analysis

    Egonex-AI/Understand-Anything

    Reads your git changes or a pull request against a prebuilt knowledge graph of the project to explain what changed, which components are affected and what is risky.

    86k GitHub starsUsed in 1 repo~1.4k tokens
    DevelopmentAuto-check passed
  • PR Design Doc

    OpenHands/OpenHands

    For a non-trivial pull request, write a self-contained HTML design doc under the temporary .pr/ directory and link a visibility-appropriate preview in the PR description, so maintainers grasp the…

    90k GitHub stars~2.4k tokensUpdated today
    DevelopmentAuto-check passed
  • WooCommerce Code Review

    woocommerce/woocommerce

    Reviews WooCommerce code changes against the project's standards, flagging backend PHP architecture, naming, documentation, data integrity and testing violations.

    11k GitHub starsUsed in 3 repos~1.1k tokens
    DevelopmentAuto-check passed

More from Dpro-at/Tel-Agent

  • Channel Extension

    Dpro-at/Tel-Agent

    A skill your agent uses when adding a new messaging channel to Tel-Agent or changing an existing one — a transport under api/channels/, its official extension manifest, its settings card, tests…

    1.1k GitHub stars~4.1k tokensUpdated 23 days ago
    Auto-check passed

Categories

Questions about Contributing

What does Contributing do?

A skill your agent uses when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking…. Contributing is an agent skill from Dpro-at/Tel-Agent. Use when contributing to Tel-Agent - picking an issue to work on, setting up the repository for the first time, starting or finishing a task, opening a pull request, or asking "what can I work on" / "how do I start" / "is my change ready to submit".

When should I use Contributing?

Contributing fits situations like: contributing to Tel-Agent - picking an issue to work on; setting up the repository for the first time; finishing a task; opening a pull request.

How do I install Contributing in Claude Code?

Run `npx skills add Dpro-at/Tel-Agent --skill contributing -a claude-code`. Or copy the skill folder (.claude/skills/contributing in Dpro-at/Tel-Agent) into .claude/skills/contributing in your project. Claude Code loads it when a task matches its description.

How do I install Contributing in Codex?

Run `npx skills add Dpro-at/Tel-Agent --skill contributing -a codex`. Or copy the skill folder (.claude/skills/contributing in Dpro-at/Tel-Agent) into .agents/skills/contributing in your project. Codex loads it when a task matches its description.

Can I use Contributing 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 Dpro-at/Tel-Agent --skill contributing -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/contributing, .gemini/skills/contributing, .github/skills/contributing and .opencode/skills/contributing in your project.

What does Contributing need to run?

Going by SKILL.md and its folder, Contributing needs the command-line tools its instructions call (gh, git, npm, node, npx and ruff). Our summary lists: Python 3; Node.js.

Does Contributing access the network?

SKILL.md names 2 domains. In commands or code: cli.github.com; the agent is likely to contact it when it follows the instructions. As links in the text: github.com. This is read from the text; nothing was executed.

Is Contributing safe to install?

Our automated static check of SKILL.md found notes only (mentions a .env file), nothing it rates as a warning. It is not a guarantee. Review the folder before installing.

What licence does Contributing use?

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

How many tokens does Contributing use?

About 3.3k tokens (SKILL.md is roughly 13k 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 Contributing?

Skills that share tags, products or a category with Contributing: Finishing a Development Branch (obra/superpowers, 296k stars), PR Babysitter (openinterpreter/openinterpreter, 69k stars), Check PR (onyx-dot-app/onyx, 32k stars) and Understand Diff Analysis (Egonex-AI/Understand-Anything, 86k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Contributing?

Dpro-at (a GitHub organization) maintains it in Dpro-at/Tel-Agent, which has 1,089 GitHub stars. The repository holds 2 skills in this directory. The repository was last updated on September 14, 2026.

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