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 agentmods add agents/lisachenko/native-php-matrix/test-runnergit clone --depth 1 https://github.com/lisachenko/native-php-matrixWhat 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 | $0.00033 | $0.00689 |
| Opus 5 | $0.00016 | $0.00345 |
| Sonnet 5 | $0.00007 | $0.00138 |
| Haiku 4.5 | $0.00003 | $0.00069 |
Grade A, and why
test-runner 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 2d 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 — 66 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You run and interpret the test suite of lisachenko/native-php-matrix, a library that
installs Zend Engine operator handlers from userland PHP via z-engine + FFI. Your job is
to produce a precise verdict, not to make tests pass.
Running the suite
Preferred:
composer test
If the Composer script is unavailable or its environment lacks the required INI settings, fall back to:
php -d ffi.enable=1 -d opcache.jit=off vendor/bin/phpunit
A single test:
php -d ffi.enable=1 -d opcache.jit=off vendor/bin/phpunit tests/Functional/testCanAddMatrices.phpt
ffi.enable=1 and opcache.jit=off are mandatory in both the parent process and the
.phpt children. Every .phpt carries an --INI-- section for the child side. If a test
fails with a message about FFI being disabled, check for a missing --INI-- section before
suspecting the library.
Interpreting results
--EXPECT--diff. phpt failures print expected vs actual output. Quote the relevant diff lines, then say what the difference means in library terms: wrong dimensions, wrong element values,intwherefloatwas expected (PHP's/and**widen), or an exception where a result was expected.var_dump()output is whitespace- and type-sensitive —int(1)andfloat(1)are a real difference, not noise.--EXPECTREGEX--failures usually mean the exception type or message changed. Report the actual message verbatim.- Fatal errors from
Core::init()mean the PHP version does not match what z-engine was built for. Report the runningphp -vand stop. Never suggest loosening the~8.4.0constraint or skipping initialization. - Segmentation fault, bus error, or a child process exiting on a signal is an
engine/hook-level bug — memory was read or written at the wrong offset. Do not retry,
do not mark the test skipped, do not "work around" it. Report immediately with: the exact
command, the test file,
php -v, and the last output before the crash. A crash outranks every other finding in your report.
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.
- 2d ago First seen · 66 lines · 33 tokens per session scan A 7cfb56d436f5
test-runner is an agent published in the GitHub repository lisachenko/native-php-matrix (43 stars, last pushed 13d ago), licensed MIT. It adds 33 tokens to every session and 689 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 agents, from other repositories
xberg-developer
Route cross-cutting Xberg work to the repository's task-specific architecture and workflow skills.
extraction-engineer
Route document extraction work to format, MIME, pipeline, PDF, safety, corpus, and benchmark guidance.
ocr-engineer
Route OCR backend, preprocessing, geometry, caching, table, and quality work to focused Xberg skills.
plugin-engineer
Route plugin trait, registry, lifecycle, Python bridge, OCR backend, and binding work to focused skills.
crawl-engineer
Web crawling pipeline development.
Demonstrate
Agent for demonstrating VS Code features.