Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .claude/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
Claude Code copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Type this inside Codex. $skill-installer <name> installs a curated skill from openai/skills. The installer writes to $CODEX_HOME/skills (default ~/.codex/skills). Restart Codex if the skill does not show up.
skills CLI
$ npx skills add simovilab/databus --skill repo-guard -a codex
Project install goes to .agents/skills/; add -g for ~/.codex/skills/.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .agents/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
Codex copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
skills CLI
$ npx skills add simovilab/databus --skill repo-guard -a cursor
Project install goes to .agents/skills/; add -g for ~/.cursor/skills/.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .cursor/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
Cursor copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
--scope user (default) or --scope workspace; --path is the subfolder of the repo that holds the skill; --consent skips the security confirmation prompt.
skills CLI
$ npx skills add simovilab/databus --skill repo-guard -a gemini-cli
Project install goes to .agents/skills/; add -g for ~/.gemini/skills/.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .gemini/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
Gemini CLI copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
GitHub CLI
$ gh skill install simovilab/databus repo-guard
Installs for Copilot at project scope by default; add --scope user for a personal install. Preview a skill first with gh skill preview. Needs GitHub CLI 2.90.0 or later (public preview).
skills CLI
$ npx skills add simovilab/databus --skill repo-guard -a github-copilot
Project install goes to .agents/skills/; add -g for ~/.copilot/skills/.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .github/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
GitHub Copilot copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
skills CLI
$ npx skills add simovilab/databus --skill repo-guard -a opencode
OpenCode documents no install command of its own. Project install goes to .agents/skills/; add -g for ~/.config/opencode/skills/.
Install the "repo-guard" agent skill from https://github.com/simovilab/databus/tree/main/.claude/skills/repo-guard into .opencode/skills/repo-guard/ in this project. Copy the whole folder (SKILL.md and every file beside it), keep the folder name "repo-guard", then confirm the skill loads.
OpenCode copies the folder itself, the same result as the manual copy. Check what it changed before you commit it.
Facts
Skill name
repo-guard
GitHub stars
101
Token cost
~1.3k tokens
SKILL.md length
626 words
Files
1
Skills in repo
1
Repo updated
First seen
Licence
Apache-2.0
At a glance
Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale.
Works in 2 steps: Automatic, zero-token — the… → On-demand, this skill — dispatches ONE…
Tasks that involve Linting and formatting
SKILL.md covers Run protocol, Conventions the agent checks…, Docs discipline and Gates (authoritative —…, plus 2 more sections
Calls git, make and uv
What it does
Repo Guard is an agent skill from simovilab/databus. Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale. Invoke before committing a feature, or when asked to "check the repo stays clean" / "guard" / "compliance-check" a diff. Complements the automatic ruff PostToolUse hook (mechanical style, zero tokens); this skill covers what a linter can't judge: docstring accuracy, doc drift, and doc updates.
Its SKILL.md is about 1.3k 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 Linting and formatting, Technical documentation and Type safety. It works with Ruff, Redis, PostgreSQL and Django. The repository describes itself as: A real-time server implementing the GTFS Schedule and Realtime specifications for transit data management. The licence is Apache-2.0.
When your agent uses it
Tasks that involve Linting and formatting
Tasks that involve Technical documentation
Tasks that involve Type safety
Example prompts
“check the repo stays clean”
“compliance-check”
“/repo-guard”
Requirements
Docker
Workflow steps
2 steps, taken from the first numbered list in SKILL.md.
1Automatic, zero-token — the ruff-guard.sh PostToolUse hook lints every edited backend/**.py on save (docstring presence D1, imports…
2On-demand, this skill — dispatches ONE diff-scoped Sonnet agent for what ruff can't judge: are docstrings accurate, are type hints…
What it can do on your machine
Read from SKILL.md and the folder at commit 1be2da1. 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:
git
make
uv
docker
pytest
From the folder's file list and the shell code blocks in SKILL.md.
Network
No URLs in SKILL.md. Its commands use git, uv and docker, which can reach the network depending on how they are called.
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
Repo Guard loads about 1.3k tokens when it runs. Until then it costs about 123 tokens; SKILL.md has 626 words of instructions outside code blocks.
Always· name and description, kept in context so the agent knows when to use it
~123
When it runs· the whole SKILL.md, loaded when a task matches
~1.3k
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 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.
Download SKILL.mdSave it as .claude/skills/repo-guard/SKILL.md (or your agent's skills folder).
name
repo-guard
description
Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale. Invoke before committing a feature, or when asked to "check the repo stays clean" / "guard" / "compliance-check" a diff. Complements the automatic ruff PostToolUse hook (mechanical style, zero tokens); this skill covers what a linter can't judge: docstring accuracy, doc drift, and doc updates.
repo-guard
Two layers keep databus release-clean:
Automatic, zero-token — the ruff-guard.sh PostToolUse hook lints every edited backend/**.py on save (docstring presence D1, imports, unused). You never invoke this; it just runs.
On-demand, this skill — dispatches ONE diff-scoped Sonnet agent for what ruff can't judge: are docstrings accurate, are type hints complete, did the change make a docs page or README stale. Run it before committing a chunk of work.
Run protocol
Scope to the diff — never scan the whole repo.bash
git -C <repo> diff --name-only main...HEAD # committed work on the branch
git -C <repo> diff --name-only # unstaged, if reviewing WIP
Keep only backend/**.py (skip migrations/, tests/, gtfs-eta/).
If nothing is in scope, say so and stop — spend nothing.
Dispatch ONE Sonnet agent (model: sonnet, general-purpose) with the brief below. Serialize if another agent is touching the git index.
Relay its result; it produces at most one style-fix commit and one docs commit.
Conventions the agent checks (the databus bar)
Area
Rule
Docstrings
One line, pep257, on every class/function. Accurate — never restates the name. ruff D1 enforces presence; you judge truth. Config ignores tests/, __init__.py (D104), nested Meta classes (D106).
Never inline redis.Redis(...). Always from databus.redis_client import create_redis_client; preserve decode_responses per call site (lifecycle guards/actions use bytes → False).
Secrets
No hardcoded tokens/passwords/keys, ever. Read from env / config().
Tests
Live in a tests/ package as test_*.py. A file named tests.py is never collected → dead.
Serializers
No duplicate field / fields="__all__" declarations.
Files
Small, focused; immutable patterns; no dead code.
Commits
Conventional (type(scope): …); one concern each; NO attribution footers; never push.
Docs discipline
When changed code touches a documented behavior, the docs must follow — docs describe as-built reality, never aspirational or stale claims.
Find affected pages cheaply: grep -rl <changed-module-or-symbol> docs/content/ backend/**/README.md — read only the hits, not the docs tree.
Update the page + the app README if either now reads wrong.
Rebuild: cd docs && uv run zensical build → must print "No issues found". docs/site/ is gitignored — never commit it.
Gates (authoritative — IDE/Pyright diagnostics are noise)
make lint — run on the host (in-container ruff can't see .gitignore, false-flags vendored gtfs-django/). Expect "All checks passed!".
make typecheck — mypy in docker; only acceptable errors are in vendored gtfs-django/.
docker compose -f compose.dev.yml run --rm orchestrator uv run pytest -q — full suite stays green; the count must not drop.
Show full SKILL.md (250 more words)Show less
Sub-agent brief (paste, substituting REPO and FILES)
You are a databus repo-cleanliness guard. Work only within these changed files: FILES (all under REPO/backend). Do NOT read the rest of the repo except the specific docs/content/** pages or README.md files that grep shows reference a symbol you changed. Read each file once.
Check each changed file against the databus conventions: docstring accuracy (one line, pep257, never merely restating the name — fix wrong/tautological ones), complete type hints, Redis via create_redis_client only, no hardcoded secrets, no tests.py (must be tests/test_*.py), no duplicate serializer declarations. Fix violations in place — behavior-neutral only; never alter logic to satisfy a linter.
Then check docs drift: for each changed module / public symbol, grep -rl <name> docs/content/ **/README.md; read only hits; update any page/README that now reads wrong so it matches the code. Rebuild docs with cd docs && uv run zensical build (must say "No issues found"; do not commit docs/site/).
Produce up to two commits, staged with explicit git add: style/fix(<scope>): … for code, docs(<scope>): … for docs. Conventional messages, NO attribution footers, do NOT push. Gates before reporting: make lint (host) clean, make typecheck only gtfs-django errors, pytest -q green with no count drop. Report commit hashes, git show --stat, gate output, and anything you fixed or found unfixable.
Efficiency contract (why this stays cheap)
Mechanical style costs zero tokens (the ruff hook, not an agent).
The agent reads only the diff + grep-matched doc pages — never the tree.
One agent, at most two commits, one pass. No speculative exploration.
Repo Guard 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.
Run this project's verification — the full local CI mirror (ruff, mypy, import contract, pytest with PostgreSQL, coverage floors, dogfood, generated docs and site, the isolated example suites, the…
Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale. Repo Guard is an agent skill from simovilab/databus. Keep the databus repo clean as it grows — verify new/changed backend code meets the project's docstring, type-hint, and structural conventions, and update any docs the change made stale.
When should I use Repo Guard?
Repo Guard fits situations like: tasks that involve Linting and formatting; tasks that involve Technical documentation; tasks that involve Type safety.
How do I install Repo Guard in Claude Code?
Run `npx skills add simovilab/databus --skill repo-guard -a claude-code`. Or copy the skill folder (.claude/skills/repo-guard in simovilab/databus) into .claude/skills/repo-guard in your project. Claude Code loads it when a task matches its description.
How do I install Repo Guard in Codex?
Run `npx skills add simovilab/databus --skill repo-guard -a codex`. Or copy the skill folder (.claude/skills/repo-guard in simovilab/databus) into .agents/skills/repo-guard in your project. Codex loads it when a task matches its description.
Can I use Repo Guard 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 simovilab/databus --skill repo-guard -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/repo-guard, .gemini/skills/repo-guard, .github/skills/repo-guard and .opencode/skills/repo-guard in your project.
What does Repo Guard need to run?
Going by SKILL.md and its folder, Repo Guard needs the command-line tools its instructions call (git, make, uv, docker and pytest). Our summary lists: Docker.
Does Repo Guard access the network?
SKILL.md contains no URLs. Its commands use git, uv and docker, which can reach the network depending on how they are called. This is read from the text; nothing was executed.
Is Repo Guard 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 Repo Guard use?
Repo Guard is published under the Apache-2.0 licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.
How many tokens does Repo Guard use?
About 1.3k tokens (SKILL.md is roughly 5.3k 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 Repo Guard?
Skills that share tags, products or a category with Repo Guard: Adk Style (google/adk-python, 22k stars), Update Dependencies (alorence/django-modern-rpc, 111 stars), Python Idioms (irahardianto/awesome-agv, 156 stars) and Python Code Style (wshobson/agents, 40k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.
Who maintains Repo Guard?
simovilab (a GitHub organization) maintains it in simovilab/databus, which has 101 GitHub stars. The repository was last updated on September 7, 2026.
Source: simovilab/databus on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.
