Mahal Systems ERP-Bench (Anchor)
Fits ERP and back-office automation agents. Borrow the parametric task generator that yields a solver-certified optimal solution and a state-based verifier per task.
What it asks the agent to do
- run a procurement workflow end to end in an ERP system
- plan and record manufacturing orders
- keep business records consistent over a long sequence of actions
What to borrow from it
grading designcase templatesandbox design
Caveats
Two-author release with the ERP product unnamed on the paper page; dataset site could not be reached to confirm license; procurement and manufacturing only.
Details
| Workloads | workflow automation, manufacturing, finance, planning |
|---|---|
| Interaction | autonomous long horizon |
| Environment | api sandbox |
| End state checked | Yes |
| Tools | ERP system interfaces |
| Grading | state diff, programmatic check |
| Size | 300 |
| Data | synthetic |
| Personal data | none |
| License | unknown |
| Harness | own repo |
| Status | active |