Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/theredbluepill/arc-interactivenpx agentmods add skills/theredbluepill/arc-interactive/check-arc-game-solvableWrote 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/theredbluepill/arc-interactive/check-arc-game-solvable)<a href="https://agentmods.dev/skills/theredbluepill/arc-interactive/check-arc-game-solvable"><img src="https://agentmods.dev/badge/skills/theredbluepill/arc-interactive/check-arc-game-solvable/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/theredbluepill/arc-interactive/check-arc-game-solvable"><img src="https://agentmods.dev/badge/skills/theredbluepill/arc-interactive/check-arc-game-solvable.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 | $0.00052 | $0.00787 |
| Opus 5.5 | $0.00021 | $0.00315 |
| Sonnet 5.5 | $0.00010 | $0.00157 |
| Haiku 4.5 | $0.00005 | $0.00079 |
Grade A, and why
check-arc-game-solvable 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 2mo 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 — 48 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Check ARC game mechanical solvability
Soul of this skill
Mechanical solvability means: for each static authored level, there exists some sequence of legal actions under the actual environment_files/ implementation that reaches the win condition (e.g. next_level(), goal state), without relying on hidden information or repo prose.
This is not the same as check-arc-game-discoverable: discoverability asks whether a cold-start participant can learn goals and rules through play. A game can be mechanically solvable yet opaque, or discoverable yet broken if a level is mathematically unwinnable. Both matter for fair ARC-style tasks.
Primary tool
From the repo root:
uv run python devtools/verify_level_solvability.py --stem <stem>
- Writes
devtools/reports/level_solvability.json(and--mdfor Markdown). proved— harness found a witness or a stem-specific proof path.counterexample— harness believes the level is unwinnable under its model (investigate; may be harness gap).tooling_gap— no strict proof in this harness (seenotes/devtools/solvers/registry.py); do not assume “pass by omission.”
Stem-specific proofs
When generic engine BFS is wrong or too large, the repo adds a dedicated strategy in devtools/solvers/ and maps the stem in devtools/solvers/registry.py.
Lights Out GF(2) (lo02, lo03, lo05): orthogonal/king neighbors with torus wrap (lo02/lo03) and the clipped closed chess-knight neighborhood (lo05, knight_clip — no wrap) are checked with GF(2) linear algebra in devtools/solvers/torus_lights_gf2.py. Initial lights_on layouts should be built by forward simulation (apply virtual clicks from all-off) so they lie in the reachable subspace — clipped kernels are rank-deficient (lo05's matrix is 56/64), so most hand-placed patterns are unreachable.
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.
- 2mo ago First seen · 48 lines · 52 tokens per session scan A 825f5ec9f55c
check-arc-game-solvable is a skill published in the GitHub repository theredbluepill/arc-interactive (57 stars, last pushed 1mo ago), licensed MIT. It adds 52 tokens to every session and 787 once invoked, about $0.0002 per session on Opus 5.5. 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
regression-suite
Map test coverage to GDD critical paths, find fixed bugs lacking regression tests, flag drift from new features.
soak-test
Soak test protocol for extended play — what to observe and log for slow leaks, fatigue, late-appearing edge cases.
playtest-report
Structured playtest report template, or turn existing playtest notes into structured feedback.
team-polish
Orchestrate the polish team — performance-analyst, technical-artist, sound-designer, qa-tester — to optimize and harden a feature or area.
develop-web-game
Use when Codex is building or iterating on a web game (HTML/JS) and needs a reliable development + testing loop: implement small changes, run a Playwright-based test script with short input bursts and intentional pauses, inspect screenshots/text, and review console errors with rendergametotext.
smoke-check
Critical-path smoke gate before QA hand-off — runs the automated suite. A failed check means the build is not QA-ready.