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 lexfrei/ccc --skill pairwisegit clone --depth 1 https://github.com/lexfrei/cccWrote 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/lexfrei/ccc/pairwise)<a href="https://agentmods.dev/skills/lexfrei/ccc/pairwise"><img src="https://agentmods.dev/badge/skills/lexfrei/ccc/pairwise/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/lexfrei/ccc/pairwise"><img src="https://agentmods.dev/badge/skills/lexfrei/ccc/pairwise.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00130 | $0.00874 |
| Opus 5.5 | $0.00052 | $0.00350 |
| Sonnet 5.5 | $0.00026 | $0.00175 |
| Haiku 4.5 | $0.00013 | $0.00087 |
Grade A, and why
pairwise 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 1mo 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 — 47 lines — stays where its author put it; the contents beside it link to each section on GitHub.
A pairwise covering array guarantees every combination of levels of every factor pair appears in at least one run, in close to the minimal number of runs. Any behavior driven by one factor or a two-factor interaction is guaranteed to fire somewhere in the array — that is the same trigger guarantee the taguchi arrays give, without their fixed 2-3-level shapes.
The tradeoff: covering arrays are not balanced. Levels appear unequal numbers of times, so averaging outcomes per level is not a fair comparison between factors. Use this skill to make the bug fire and to shortlist suspects; rank and confirm with balanced follow-ups.
Step 1 — factors and levels
Same rules as the taguchi skill: factors must be independently settable, every level discrete and articulable, observable-but-unsettable factors recorded as covariates. Level counts are unrestricted — that is the point of this skill.
Step 2 — generate the run sheet
Run the bundled generator (relative to this skill's base directory):
python3 scripts/pairwise.py "net=fast,slow,flaky,off" "browser=chrome,firefox,safari" "os=linux,macos"
It prints a markdown run sheet plus the run count against the full factorial. The output is self-verified: the script crashes rather than print an array with an uncovered pair. It restarts the greedy search from 64 seeds and keeps the shortest array — the seeds cost milliseconds, the runs they save cost CI round-trips (--restarts trades one for the other, --json emits the rows for a driver script).
Combinations that cannot be run go in as constraints, not as rows you quietly skip. Safari does not run on Linux; version X does not build against Y. Dropping such a row by hand takes its pairs down with it and the coverage guarantee is gone without a word:
python3 scripts/pairwise.py "browser=chrome,firefox,safari" "os=linux,macos" "net=fast,slow" \
--exclude "browser=safari & os=linux"
Excluded rows are never generated, pairs that no legal row can contain leave the target set, and everything still reachable stays covered.
What ships with it
2 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.
- 1mo ago First seen · 47 lines · 130 tokens per session scan A 510014c4b0a9
pairwise is a skill published in the GitHub repository lexfrei/ccc (11 stars, last pushed 14d ago), licensed BSD-3-Clause. It adds 130 tokens to every session and 874 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 2026-08-31.
Other skills, from other repositories
research-engineer
An uncompromising Academic Research Engineer. Operates with absolute scientific rigor, objective criticism, and zero flair. Focuses on theoretical correctness, formal verification, and optimal implementation across any required technology.
eval-dataset-design
A guide to creating reliable test sets for AI agents. Each evaluation task includes the request, the simulated user's behaviour, a reset starting state, and a success check that can be independently verified.
tika-eval-compare
Compare extracts from two Tika builds over a corpus to detect regressions in content, encoding, exceptions, and embedded-document handling. Use for "compare before/after extracts", "eval this change against the corpus".
cua-driver
Use Cua Driver for desktop or browser tasks that are awkward or unavailable through Bash/APIs, or when the user explicitly wants GUI interaction: app testing, visual bug reproduction, form filling, calendar entry, screenshots, and demo recording. Also covers Cua setup; not OpenAI Codex Computer Use or web research.
jetson-validate-image
Use after jetson-flash-image to run static BSP checks, on-target smoke/regression tests on a flashed DUT, or both. Not for build or flash steps. Triggers: validate bsp, on-target validation.
atmos-validation
Validate Atmos projects, components, arbitrary JSON Schema inputs, EditorConfig, and GitHub Actions; use affected-file selection and native CI annotations.