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 OutlineDriven/outline-driven-development --skill carbon-langgit clone --depth 1 https://github.com/OutlineDriven/outline-driven-developmentWrote 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/outlinedriven/outline-driven-development/carbon-lang)<a href="https://agentmods.dev/skills/outlinedriven/outline-driven-development/carbon-lang"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/carbon-lang/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/outlinedriven/outline-driven-development/carbon-lang"><img src="https://agentmods.dev/badge/skills/outlinedriven/outline-driven-development/carbon-lang.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 | $0.00046 | $0.01643 |
| Opus 5 | $0.00023 | $0.00822 |
| Sonnet 5 | $0.00009 | $0.00329 |
| Haiku 4.5 | $0.00005 | $0.00164 |
Grade A, and why
carbon-lang 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 — 69 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Carbon
Carbon is an experimental successor language to C++ from the carbon-language project. As of 2026-09-05 there is no 1.0 and no 0.1: the 0.1 milestone is the minimum language for evaluation, and the 2025 roadmap revision moved its earliest realistic date to the end of 2026 after memory-safety design was added to its scope. The only downloads are nightly toolchain tarballs tagged v0.0.0-0.nightly.YYYY.MM.DD, on limited platforms. Treat every finding below as dated to the nightly you ran.
Contract
| Field | Bound contract |
|---|---|
| Trigger | A team with a large C++ code base asks whether to evaluate Carbon, wants to run the toolchain, wants to call C++ from Carbon, or asks how Carbon's safety model and readiness compare with C++. |
| Authority | Reversible local: writes only the toolchain checkout or extracted nightly tarball, Carbon source files, and --output artifacts in the working directory; rollback is deleting them. No remote mutation. |
| Side effect | A Bazel build of the toolchain compiles a pinned LLVM commit and takes substantial disk and time. |
| Done | The evaluation report states the nightly date or commit used, what compiled and what did not, and a decision (evaluate further or stay on C++) tied to the constraints in step 7. |
Inputs
- The C++ code base under consideration and its constraints: stdlib and Boost dependence, template depth, platforms.
- A toolchain source: a nightly tarball from the GitHub releases page, or a checkout of
https://github.com/carbon-language/carbon-langwith Clang 19 or newer,libc++-dev,libc++abi-dev, andlldinstalled. Bazel is pinned by.bazelversion(8.6.0 at grounding) and fetched byscripts/run_bazelisk.py; LLVM is pinned as an upstream commit throughgit_overrideinMODULE.bazel, not as a release number. - The evaluation question: syntax and ergonomics, C++ interop, or readiness.
Procedure
- Record the toolchain identity. From a nightly: extract the tarball, take the version from the
carbon_toolchain-${VERSION}directory name, and run./carbon_toolchain-${VERSION}/bin/carbon help. From source:./scripts/run_bazelisk.py run //toolchain -- helpbuilds and runs the driver (target//toolchain:carbon, alias//toolchain:toolchain);//toolchain/install:carbon_toolchainpackages the tarball. Done when: the nightly tag or commit hash is written into the report. - Compile the README example to prove the toolchain works:
What ships with it
1 file 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.
- 2d ago First seen · 69 lines · 46 tokens per session scan A 83e77f96d1e1
carbon-lang is a skill published in the GitHub repository OutlineDriven/outline-driven-development (52 stars, last pushed 2d ago), licensed Apache-2.0. It adds 46 tokens to every session and 1,643 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 2026-09-06.
Other skills, from other repositories
bounded-model-checking-c
Use when C or C++ code needs memory-safety or undefined-behavior guarantees proved with CBMC, or ACSL contracts checked with Frama-C Eva or WP. Not for choosing the proof policy: use proof-driven.
cpp-modules
C++20 modules skill for modern C++ projects. Use when working with named modules, module partitions, header units, CMake MODULESOURCES, Clang -fmodules-ts, BMI caching issues, or migrating from headers to modules. Activates on queries about C++20 modules, import statements, module interface units, header units, or BMI…
openmp
OpenMP skill for shared-memory parallel programming. Use when writing parallel for loops, reductions, task parallelism, SIMD directives, GPU offloading, or profiling with Score-P/TAU. Activates on queries about OpenMP, pragma omp, schedule static dynamic, reduction, false sharing, or OMPNUMTHREADS.
compiler-optimizations-deep
Deep compiler optimizations skill for RA, ISel, and PGO. Use when explaining register allocation, instruction selection, LICM, vectorization limits, or profile-guided optimization beyond -O3. Activates on queries about register allocation, instruction selection, LICM, auto-vectorization failure, PGO, or BOLT.
xmake-targets
Use when configuring targets in xmake.lua — setting kind (binary/static/shared), adding files/includes/links, defining dependencies between targets, and applying per-target compile/link flags.
cpp
Use when writing modern C++ (17/20/23). Covers RAII, smart-pointer ownership, move semantics, ranges, concepts, and eliminating undefined behavior with sanitizers.