internationalcolorconsortium/iccdev/valgrind-analysis
Claude Code and agent skill by internationalcolorconsortium/iccdev
- community
Checking that the SKILL.md file is still in the repository.
Build and run iccDEV tool lanes under Memcheck, Helgrind, DRD, Massif, or Callgrind while preserving analyzer evidence.
Install
npx vskill@latest install internationalcolorconsortium/iccdev/valgrind-analysisSKILL.md
Valgrind Analysis
Use this skill for uninitialized-memory, invalid-access, leak, synchronization, heap-growth, or call-profile investigations against iccDEV tools.
Workflow
- Read
../../../docs/valgrind-analysis.md. - Select the smallest registered target that reaches the affected path.
- Run
.github/ci/valgrind/build.sh --target TARGETto produce a separate
non-sanitized Debug build.
- Confirm the build rejects cached sanitizer flags and sanitizer-instrumented
binaries, including statically linked runtimes.
- For runner or image QA, run
.github/ci/valgrind/self-test.shafter building
the dump target.
- Run
.github/ci/valgrind/run.sh --tool TOOL TARGETwith a bounded timeout. - Preserve the generated
summary.tsv, analyzer log, stdout, stderr, and
generated work files.
- For concurrency, compare Helgrind and DRD rather than treating either as a
substitute for TSan. Confirm that the selected input reaches the worker threshold before claiming that a requested thread count exercised concurrent workers.
- For issue #2592 or
m_jobReadynotification changes, run
.github/ci/valgrind/issue-2592-regression.sh; require both analyzers to be clean and their generated TIFFs to match.
- Attribute findings to stack frames and observed synchronization behavior,
not input filenames.
- For
.icccorpus sweeps, buildpawgordumpand usecorpus.sh; do not
use word-splitting command substitution around find.
- Report the exact source revision, command, exit code, error count, and
evidence directory.
Use --allow-findings only when collecting a known or deliberately compared finding. Never use it to turn an unexplained result into a clean gate.
Container path
The unified image exposes iccdev-valgrind-build, iccdev-valgrind-run, iccdev-valgrind-status, iccdev-valgrind-validate, and iccdev-valgrind-self-test. These commands write to /workspace/valgrind by default and must not wrap /workspace/build.
The repository's .github/ci/valgrind/corpus.sh supports mounted corpus paths inside the unified image. Preserve per-profile evidence outside the container; prefer tar streaming when a writable bind mount has a different owner.
References
../../../docs/valgrind-analysis.md../../prompts/valgrind-analysis.prompt.md../sanitizer-repro/SKILL.md../regression-container-maintainer/SKILL.md
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryChecking.github/skills/valgrind-analysis/SKILL.md
Popularity
- 7-day trend score
- 5
Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/internationalcolorconsortium/iccdev/valgrind-analysis)