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 chen3feng/agent-skills --skill detect-tool-vendor-by-querygit clone --depth 1 https://github.com/chen3feng/agent-skillsWrote 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/chen3feng/agent-skills/detect-tool-vendor-by-query)<a href="https://agentmods.dev/skills/chen3feng/agent-skills/detect-tool-vendor-by-query"><img src="https://agentmods.dev/badge/skills/chen3feng/agent-skills/detect-tool-vendor-by-query/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/chen3feng/agent-skills/detect-tool-vendor-by-query"><img src="https://agentmods.dev/badge/skills/chen3feng/agent-skills/detect-tool-vendor-by-query.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 | $0.00032 | $0.01682 |
| Opus 5 | $0.00016 | $0.00841 |
| Sonnet 5 | $0.00006 | $0.00336 |
| Haiku 4.5 | $0.00003 | $0.00168 |
Grade A, and why
detect-tool-vendor-by-query 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 10d 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 — 151 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Detect tool vendor by query, not by name
When to use
You are writing code that branches on which compiler, linker,
interpreter, or other build tool is in use — GCC vs. Clang, CPython
vs. PyPy, BSD vs. GNU coreutils, system Python vs. a pyenv shim — and
your current approach is to look at the executable's name or path
("gcc" in argv0, which python3, basename $CC).
Signal phrases: "we assume /usr/bin/gcc is GCC", "the driver is
called cc so it must be…", "python3 on PATH is 3.x", or a bug
report that reproduces only on macOS / only on a system with a
vendor-renamed toolchain.
Problem
Tool names lie. Common traps:
- macOS ships
/usr/bin/gccbut it's Apple Clang under the hood. Any substring check like'gcc' in cc_pathmisclassifies it as GNU and then tries to pass GNU-only flags (-static-libgcc,-flto=auto,-fno-fat-lto-objects, …), which Clang rejects at link time. ccis a symlink whose target varies by distro: GCC on most Linux, Clang on FreeBSD / macOS, tcc on some minimal installs. The namecctells you nothing.python3on PATH may be the system's frozen 3.9, a pyenv shim, a Homebrew keg, or a virtualenv. Checking the path (/usr/bin/python3→ "system Python") gives you the source, not the version.- GNU vs BSD coreutils —
sed,awk,tar,dateon macOS are BSD variants with different flag sets; they're spelled exactly the same as their GNU cousins. - Cross-compilers and wrappers (
ccache gcc,distcc clang,arm-linux-gnueabihf-gcc) contain the vendor substring but may forward to a different backend than the name implies.
The fix is always the same: ask the tool.
Solution
- Run the tool's identity query once, at init time, and cache the result. Don't re-probe on every call site.
- Parse a known-stable field, not the whole banner:
- C/C++ drivers: first line of
cc --version. Look for'Apple clang'/'clang'/'gcc'/'Free Software Foundation'. Normalize to a closed enum ({'clang', 'gcc', 'unknown'}). - Python:
python -c 'import sys; print(sys.version_info[:2])'orpython -Vand parse"Python X.Y.Z". Compare as a tuple, never as a float (3.10 < 3.9if you use floats). - coreutils:
sed --version 2>&1 | head -1→"GNU sed"vs. "invalid option" on BSD.
- C/C++ drivers: first line of
- Have a defined
unknownbucket. Exotic toolchains exist; don't let the detector crash on them. Emit a warning and fall back to the most conservative flag set. - Gate feature flags on the cached vendor, not on the path:
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.
- 10d ago First seen · 151 lines · 32 tokens per session scan A eaae4350c03b
detect-tool-vendor-by-query is a skill published in the GitHub repository chen3feng/agent-skills (5 stars, last pushed 4mo ago), licensed Apache-2.0. It adds 32 tokens to every session and 1,682 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 2026-08-31.
Other skills, from other repositories
sentry-app-setup
Create and configure a Sentry internal integration so the assistant can manage issues, alerts, and releases under its own identity.
burrow-system-tools
Diagnose and fix the user's Mac with Burrow's local MCP tools (burrowdoctor, burrowsnapshot, burrowtopprocesses, burrowprocessusage, burrowports, burrowanalyze, burrowdiskforecast, burrowdupes, burrowanomalies, burrowagentaudit, burrowclean, …). Use whenever the Mac is slow, hot, loud, low on disk, draining battery…
wispterm-diagnostics
Use when a user wants to report, troubleshoot, or collect context for a WispTerm issue, including crashes, rendering/DPI glitches, high CPU, keyboard/input bugs, selection/copy/scrolling, SSH/SCP failures, SSH image preview failures, HTML preview/browser panel failures, SSH disconnects such as…
self-diagnose
Sutando introspection — read logs + git + memory + build log for a chosen time window and produce a concise narrative of what the agent has been doing, what's broken, and what to prioritize next.
call-diagnostics
Analyze phone call observability data, detect problems, track them across calls, and recommend systematic repairs.
regression-search
Search phone-call history for when a feature regressed (find-regression.py) and drill into a single call to see what went wrong (diagnose-call.py). Skips reading 100+ transcripts by hand.