# Coverage

> Export source coverage of test

- Skill: `firedancer-io/coverage` (Agent Skill)
- Install (CLI): `npx skillmds@latest add firedancer-io/coverage`
- Raw SKILL.md: https://api.skillmd.com/api/skills/firedancer-io/coverage/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: firedancer-io (https://skillmd.com/u/firedancer-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/firedancer-io/coverage

---


Export source coverage of unit test $0:

1. Build: `make -j CC=clang EXTRAS=llvm-cov BUILDDIR=clang-cov $0`
2. Run: `CLANKER=1 ./contrib/test/single_test_cov.sh build/clang-cov/unit-test/$0`
   - produces `default.profdata`
3. Export
   - Per file: `llvm-cov show --instr-profile default.profdata build/clang-cov/unit-test/$0 -sources <src/foo/bar.c> -show-instantiations=false`
   - Per function: `llvm-cov show --instr-profile default.profdata build/clang-cov/unit-test/$0 --name=<function_name>`
   - All files overview (verbose!) `llvm-cov report --instr-profile default.profdata build/clang-cov/unit-test/$0`

If no further context/instructions are given, attempt to find under-covered areas of code:
Skim the test code to understand what it's trying to test, then drill down the cov report to see if it actually covers that code.