Tool-call failures: retry-strategy comparison
Experiment design set, results await raw records
This page gives only the experiment design, not the results. Numbers we haven’t run are never written; paste raw records into the table after running, then write the conclusion.
Question to answer
Inject tool-call failures on purpose (rate limit / timeout / malformed), how do different retry strategies (none / fixed-interval / backoff) affect task success rate and time?
Control groups
No retry / fixed-interval retry / exponential backoff; 3 runs each.
Environment & version
- Host version
- To fill
- Model ID
- To fill
- Mode / permissions
- Keep both groups identical
- Input files
- Record hashes so both runs use the same files
Fields to log
| # | Field | Result |
|---|---|---|
| 1 | Host and model | to be tested |
| 2 | Failure injection method | to be tested |
| 3 | Task success rate | to be tested |
| 4 | Total wall-clock time | to be tested |
| 5 | Extra token cost | to be tested |
Conclusion
To be tested. Any conclusion must also state what it can't generalize to.
Why we don’t write results yet
Backoff is usually steadier but slower; this site records the trade-off, no blind retry advice.