What is upskill AGENTS.md?

Repository instructions for a Python project, including its supported Python version, coding rules, and required local checks. Ruff is a tool for formatting and linting Python code, while type checking finds annotation-related errors.

Use them when modifying the huggingface/upskill codebase, especially before running formatting, linting, type checking, or tests.

They give the coding agent a clear quality baseline and the same checks used by continuous integration, or CI, which automatically tests changes.

upskill AGENTS.md — Instructions file by huggingface

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods

npx agentmods add instructions/huggingface/upskill/agents-md

Clone the repo

git clone --depth 1 https://github.com/huggingface/upskill

Made for: Codex, OpenCode.

Per session 443 This file is loaded in full into every session.

When invoked 443 The same file — it is already loaded in full.

Security scan A 0 findings. Scan, not verified.

Origin original No closer match found in the catalogue.

Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

Model	Per session	Once invoked
Fable 5	$0.00443	$0.00443
Opus 5	$0.00221	$0.00221
Sonnet 5	$0.00089	$0.00089
Haiku 4.5	$0.00044	$0.00044

Measured yesterday against content hash f3072b61d038, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

upskill AGENTS.md scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

AGENTS.md · 62 lines

What it actually says

Agent guidance

Code quality baseline

Target Python 3.13.
Keep changes compatible with the repo's pyproject.toml settings.
Prefer small, typed, composable functions over large command-style blocks.
Avoid introducing new complexity suppressions unless there is a strong reason.

Required local checks

Before finishing a change, run:

uv sync --extra dev
uv run scripts/format.py
uv run scripts/lint.py
uv run scripts/typecheck.py
uv run --extra dev pytest -v

CI enforces the same flow in .github/workflows/ci.yml.

Ruff rules

Formatting and linting are enforced with Ruff.

Line length: 100
Target version: py313
Enabled lint families:
- B - bugbear
- C90 - cyclomatic complexity
- E - pycodestyle errors
- F - pyflakes
- I - import sorting
- RUF - Ruff-specific rules
- SIM - simplifications
- TCH - type-checking import hygiene
- UP - pyupgrade
E501 is ignored; let ruff format own line wrapping.
Cyclomatic complexity limit: 15

Type checking

ty is required for src, tests, and scripts.
Add or improve annotations when touching code that is ambiguous to the type checker.
Prefer explicit protocols / typed helper structures over object when wiring dynamic APIs.
Keep type-only imports behind TYPE_CHECKING when appropriate.

Tests

Add or update tests for behavior changes.
Keep the test suite passing with pytest.
Use focused unit-style tests for logic changes when possible.

Practical authoring guidance

Prefer refactoring over adding broad ignores.
If a function is nearing the complexity limit, split it before adding more branches.
Keep CLI orchestration, model resolution, and persistence logic separated when possible.
When adding developer tooling, update the README and CI together.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

yesterday First seen · 62 lines · 443 tokens per session scan A f3072b61d038

Subscribe to this mod's changes

upskill AGENTS.md is an instructions file published in the GitHub repository huggingface/upskill (738 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 443 tokens to every session, about $0.0022 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Code review Languages Testing

codex AGENTS.md

AGENTS.md instructions for openai/codex, covering rust/codex-rs, the codex-core crate, code review rules, crate api surface and model visible context.

openai/codex · 5,182 tokens

buildNext

Working notes and architecture documentation for the new esbuild-based build system in build/next. Use when making changes to the new build pipeline (transpile/bundle commands, NLS plugin, source-map handling, resource copying, or self-hosting watch tasks).

microsoft/vscode · 6,785 tokens

upskill AGENTS.md