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 kouroshez/coding-os --skill motion-reviewgit clone --depth 1 https://github.com/kouroshez/coding-osWrote 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/kouroshez/coding-os/motion-review)<a href="https://agentmods.dev/skills/kouroshez/coding-os/motion-review"><img src="https://agentmods.dev/badge/skills/kouroshez/coding-os/motion-review/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/kouroshez/coding-os/motion-review"><img src="https://agentmods.dev/badge/skills/kouroshez/coding-os/motion-review.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 | $0.00115 | $0.01314 |
| Opus 5.5 | $0.00046 | $0.00526 |
| Sonnet 5.5 | $0.00023 | $0.00263 |
| Haiku 4.5 | $0.00012 | $0.00131 |
Grade A, and why
motion-review 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 28d 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 — 101 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A recording is useless until its frames are in your context. This skill turns one into the small set of frames worth looking at, then tells you what to judge them for.
The one fact that drives everything
A UI recording is a step function, not a film. Measured on real captures: ~84% of the timeline is a settled state and ~16% is transition — but ~30% of the frames sit inside those short transitions, because screen recorders are variable-rate and emit a frame only when the picture changes. Every difference-based frame selector therefore spends the budget backwards, and every cut-based one returns nothing at all.
So: one frame per settled state, N frames per transition. Why each ffmpeg built-in fails, with numbers: references/frame-selection.md.
Process
-
Get a recording. Five sources, one output — a local video file. Exact commands and the trap each one carries: references/capture-sources.md. If you are driving the app yourself, pick a rung and its claim tier first: references/driving-ladder.md.
-
Select the frames.
python3 scripts/frames.py <recording> --out <frames-dir>Stdlib only; needs
ffmpeg/ffprobeon PATH and nothing else. It prints a frame list, a per-transition timing profile, and a role label for every frame.--jsonfor the machine-readable form. Preflight the host withpython3 scripts/doctor.py. -
Read every frame it lists. In one message, parallel calls. The labels are the map:
settled state 2 of 4is judged for layout, type, contrast and state correctness;transition 3 of 3, frame 4 of 7, t=+244msis judged for trajectory — what moves, what stays, whether it overshoots. -
Judge against the contract, never "my change landed": assets/review-contract.md. It carries the defect classes a still frame structurally cannot show, and the finding grammar.
What ships with it
8 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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.
- 28d ago First seen · 101 lines · 115 tokens per session scan A 39255c3846aa
motion-review is a skill published in the GitHub repository kouroshez/coding-os (6 stars, last pushed 5d ago), licensed Apache-2.0. It adds 115 tokens to every session and 1,314 once invoked, about $0.0005 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 2026-09-20.
Other skills, from other repositories
potpie-cli
Use when the task is centered on running, explaining, configuring, or troubleshooting the potpie command: doctor, login, pot management, source registration, resolve/search/record, graph workbench reads/writes, resource (document payload) commands, and pot scope behavior.
potpie-debug-memory
Use while debugging or troubleshooting failures, flaky tests, incidents, production alerts, CI failures, local dev setup issues, repeated bugs, prior fixes, failed attempts, and verification history.
analyze-type-infer
A compiler analysis step that infers a symbol's type from its assignments. A type describes what kind of value something holds, such as a number, text, or Boolean value.
check-name-real
A name checker that verifies every required symbol has been declared. A symbol is a named program entity such as a variable or function.
compiler-0355bffb
A virtual-machine debugging skill for executing one instruction at a time. A virtual machine is a program that runs a compact instruction format instead of native machine code.
compiler-2f8c8f39
A call-graph skill that maps which functions call which other functions. A call graph is a network whose nodes are functions and whose links represent calls.