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.
npx skills add kitao/pyxel-skill --skill pyxel-skillgit clone --depth 1 https://github.com/kitao/pyxel-skillWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/kitao/pyxel-skill/pyxel-skill)<a href="https://agentmods.dev/skills/kitao/pyxel-skill/pyxel-skill"><img src="https://agentmods.dev/badge/skills/kitao/pyxel-skill/pyxel-skill/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/kitao/pyxel-skill/pyxel-skill"><img src="https://agentmods.dev/badge/skills/kitao/pyxel-skill/pyxel-skill.svg" alt="Reviewed on agentmods" width="80" height="20"></a>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 5.1 | $0.00042 | $0.00737 |
| Opus 5 | $0.00021 | $0.00368 |
| Sonnet 5 | $0.00008 | $0.00147 |
| Haiku 4.5 | $0.00004 | $0.00074 |
Grade A, and why
pyxel 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 13d ago.
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.
How it starts
The opening of the file, as written. The whole thing — 56 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Pyxel
Build the smallest complete game that satisfies the request, then verify it from observed behavior. Keep the process proportional to the task.
Runtime
Expect the pyxel MCP namespace from pyxel-mcp 1.2+ with Pyxel 2.9.6+ and Python 3.11+. Its eight tools are observation verbs:
validatereports syntax errors and recognizable code patterns.rundrives frames and captures state, screen images, screen grids, or video.pyxel_inforeports versions, examples, and resource URIs.read_palette,read_image,read_tilemap, andread_audioinspect the corresponding Pyxel data.diff_framescompares two captured PNGs.
If pyxel-mcp is unavailable, register it with uvx pyxel-mcp install; if older than 1.2.0, update it with uvx --refresh-package pyxel-mcp pyxel-mcp install. While blocked, use focused logic tests plus direct Pyxel headless runs temporarily, and state that visual and interaction verification is weaker.
Workflow
- Infer the smallest playable scope. Ask only when a missing choice would materially change the game.
- Implement a complete slice: entry state, controls, objective, and a retry or terminal state when the genre needs one.
- Run
validate; fix errors and review relevant warnings. Readpyxel://validation-patternsonly when a category needs explanation. - Use
runfrom frame 0 with deterministic input. Capturestatefor mechanics andscreen_imagefor what the player sees. Preferuntilwith snapshots at"end"for event-driven checks. - Read the run result's log field (
log) even whenokis true. Inspect the captured image for the task-specific result; nonblank output alone is not evidence of a correct scene. - Iterate only on observed defects. Add focused logic tests when rules are easier to prove outside rendering.
- Report controls, changed files, and exact verification results.
Minimum Evidence
validatehas no syntax errors; relevant warnings are resolved or explained.- A smoke
runreaches its intended stop. - At least one captured frame is inspected directly.
- At least one task-specific state predicate is checked.
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.
- 13d ago First seen · 56 lines · 42 tokens per session scan A 4563b35dfec4
pyxel is a skill published in the GitHub repository kitao/pyxel-skill (12 stars, last pushed 1mo ago), licensed MIT. It adds 42 tokens to every session and 737 once invoked, about $0.0002 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.
Other skills, from other repositories
pygame-core
Structure a pygame (pygame-ce) game in Python: the init/event/update/draw loop, delta-time movement, Surface/Rect blitting, keyboard/mouse input, and Sprite/Group management with collision. Use when building or debugging a pygame game — when the user mentions pygame, pygame-ce, the game loop, blit, Surface, Rect…
vibeue
Unreal Engine 5 development using the VibeUE Python API. Use when working in Unreal Engine — blueprints, state trees, materials, actors, landscapes, animation, niagara, widgets, sound, foliage, gameplay tags, enhanced input, skeletons, PCG (procedural content generation), and more. VibeUE is an extension of Unreal's…
python-lsp
Semantic Python code queries via a stdio LSP client driving pyright-langserver. Provides binding-resolved go-to-definition, find-references, hover types, type diagnostics, file symbol outlines, and project-wide symbol search — name resolution and type inference that tree-sitter and ripgrep cannot do. Use when you need…
searching-codebases
Binding-resolved Python symbol queries via pyright — every true caller (--refs), the real definition (--def), or an inferred signature (--hover) of a .py symbol, excluding the same-named false positives text search cannot tell apart. Use when a task needs ALL callers or users of a named Python symbol and grep would…
demo-1-terminal-arcade
Demo skill that builds a colorful, playable terminal Snake game in a single pure-stdlib Python curses file — score HUD, speed ramp, eat/death animations, and a game-over screen.
HA Integration Dev
Home Assistant custom integration development in Python. Covers customcomponents, DataUpdateCoordinator, configflow, OAuth2, conversation agent, HACS publishing, device registry, entity platforms, services, repair issues, diagnostics, Bluetooth integrations, and multi-coordinator patterns.