Individual skill
mpi
Use when writing or debugging MPI programs: sends and receives, collectives, non-blocking requests, communicators, MPI-IO, hybrids with OpenMP, or mpirun launches. Not for threads alone: use openmp.
.devin/skills/mpi/SKILL.md
Source commit b0e8ce89a19fac880251dc3ea1babfeb4503a4fe
What was checked
- Matches the observed repository HEAD
- Yes
- Scanner rules are current
- No
- Repository scan complete
- No
- Skill scan complete
- No
- Observed behavior
- Not checked yet
- Last checked
- Nov 4, 2026
These are observations at the recorded commit, not a guarantee of safety. Review the source before running it.
Install this skill
Choose an agent in your Stack or copy a command below. Commands fetch the pinned commit and copy only this skill directory. Existing destinations are refused. Copying a command does not install anything.
An install command is unavailable: current rules, complete repository and skill scans, an eligible behavior level and a permitted license are required.
Repository context
- Repository stars (not skill ratings)
- 55
- Repository last updated
- Sep 28, 2026
- License
- Apache-2.0
Observed agent integration: No specific integration evidence. Standard SKILL.md installation does not imply verified compatibility.
- odd Repository topic
- odin Repository topic
- outline Repository topic
- outlines Repository topic
- coding-paradigm Repository topic
- outline-driven-development Repository topic
- claude-plugins Repository topic
- codex-cli Repository topic
- gemini-cli-extensions Repository topic
- outline-driven Repository topic
- writing Derived category
- testing Derived category
Files and excerpt
File metadata and permitted, redacted excerpts come from the recorded scan. Source links are pinned to that commit.
Browse files
SKILL.mdcore · 8,284 bytesSKILL.mdcore · 8,284 bytesSKILL.mdcore · 8,284 bytesagents/openai.yamlasset · 220 bytes
Read excerpt
# MPI
MPI is the message-passing standard for distributed-memory parallel programs: one process per rank, explicit sends, receives, and collectives over a communicator. The commands below are exercised against Open MPI 5.0.10, the implementation on the grounding host; MPICH and vendor MPIs accept the same API and a different `mpirun` option set.
## Contract
| Field | Bound contract |
|---|---|
| Trigger | The user parallelizes across nodes or sockets, implements a distributed algorithm, combines ranks with OpenMP threads, runs under a scheduler, debugs a hang or a truncation error, or needs parallel file output. |
| Authority | Reversible local: writes only MPI source files, compiled binaries, hostfiles, and output data files in the working directory; rollback is deleting them. Scheduler submissions and remote launches are proposed to the user. No remote mutation. |
| Side effect | Source, binaries, and data files on disk. `mpirun` on the local host starts one process per rank. |
| Done | The program compiles with `mpicc`, runs under `mpirun -np N` with the expected output on every rank, every send has a matching receive, every non-blocking request is completed, and the scaling issue named in the request has a cause and a fix. |
## Inputs