Foam-Agent · Agentic Research Directory

Foam-Agent

Foam-Agent plans, runs and repairs OpenFOAM workflows using solver feedback. In the v3 author evaluation, 97 of 110 Basic cases executed, while only 46 met the stricter execution-and-fidelity criterion. This distinction matters: runnable simulations can still be inaccurate. Public code is available; industrial and experimental validation remain outside these results.

L3Multi-step tool-using agentic workflow Source Peer-reviewed source Outcome Author-reported

Evidence & verification

Source level

Peer-reviewed source
Grounded in a peer-reviewed publication. This describes the source; it does not mean the reported results have been independently reproduced.

Outcome verification

Author-reported
Outcomes are reported by the team that built the system. The entry is source-checked against the team's own publications; not independently reproduced.

Sources checked

2026-10-19 (UTC)

Checked source version: 2505.04997v3. Claims refer to this manuscript snapshot; final journal/proceedings version parity was not verified. Author-reported; independent replication not established in this review. Publication metadata confirms ICML 2024 proceedings (SGA) or CMAME 461:119271 (Foam-Agent); the peer-reviewed-work label does not assert final-text parity.

Source-checked does not mean independently reproduced. See the Method page for how these labels are assigned.

Sources (2)

Research applicability

Documented engineering research workflow, with source-version and evaluation limits retained.

Getting started

Read the primary manuscript and its limitations first. Check implementation access and terms separately; this review did not run an installation, API call to a research system, or experiment.

Maintenance

Source snapshot checked 2026-10-19. Current availability and experimental version parity are separate claims.

License

Research code/data/manuscript terms require separate verification; repository metadata alone is not a reuse clearance.

Workflow & human oversight

L3 applies only to the user-requested OpenFOAM configuration and repair workflow. A user supplies simulation requirements; the author evaluation checks outputs against expert reference fields. Per-iteration human approval is not claimed. Planning generates files; solver logs drive bounded correction. Execution success is separated from field-level fidelity. The loop stops at execution success or its configured iteration cap. Successful execution alone is not physical correctness or unrestricted scientific autonomy.

Workflow level is not an evidence score; higher is not better evidence.

Approach

Architect plans configuration files; Runner executes OpenFOAM; Reviewer diagnoses logs, edits files and repeats under an iteration cap.

Outcomes (as reported by the originating team)

Claude 3.5 Sonnet: 97/110 Basic cases execute (88.2%); strict execution-plus-top-fidelity success is 46/110 (41.8%). Fidelity buckets are 46/13/51; 13 execution failures and 38 execution-pass/fidelity-fail cases. Advanced execution: 10/16 (62.5%).

Limitations

Crash-free CFD is not scientific correctness; tutorial-derived evaluation and solver/model errors limit transfer. Current README or other-model rates are not substituted for v3 results.

Reproducibility

Author-reported; independent replication not established in this review. Documentary review only; no candidate code, model, simulation or experiment was executed. A public repository or result artifact is not proof of replication.

Adoption requirements

Untested suggestion: inspect the pinned sources and validate data, evaluator assumptions, implementation availability and researcher controls before planning a bounded trial. No setup or trial was executed.

Comments

No comments yet.

Add a comment

Comments are held for moderation before appearing. Critical or corrective comments are welcome — especially evidence that an entry overstates what has been verified.