Can you trust Claude Sonnet 5.5 for space work?
By Anthony Vincent Caracappa, Founder and Self-Directed Space Researcher, Viventine Space Systems · updated
On its own, trust it for baseline knowledge. Connected to Orbit Sentinel, trust it for catalog lookup, constellation & fleet status, debris & conjunction context, regulatory & licensing, company & market intelligence, ground infrastructure, baseline knowledge.
| Job | AloneClosed-book | With dataWith Orbit Sentinel tools | n | ||||
|---|---|---|---|---|---|---|---|
| right / wrong | range | declined | right / wrong | range | declined | ||
| Catalog lookup | 60 / 11 | 45–73 | 29 | 100 / 0 | 92–100 | 0 | 45×3 |
| Constellation & fleet status | 42 / 7 | 28–57 | 51 | 99 / 1 | 90–100 | 0 | 43×3 |
| Debris & conjunction context | 40 / 21 | 27–56 | 38 | 96 / 2 | 85–99 | 2 | 42×3 |
| Regulatory & licensing | 31 / 8 | 19–46 | 62 | 93 / 4 | 80–98 | 3 | 38×3 |
| Company & market intelligence | 37 / 24 | 24–52 | 39 | 99 / 0 | 90–100 | 1 | 41×3 |
| Ground infrastructure | 28 / 18 | 16–43 | 55 | 96 / 4 | 85–99 | 0 | 40×3 |
| Baseline knowledge | 100 / 0 | 87–100 | 0 | 99 / 1 | 84–100 | 0 | 25×3 |
| Orbit Sentinel-only questions | 8 / 4 | 2–25 | 88 | 99 / 0 | 84–100 | 1 | 25×3 |
% right / % wrong per cell; color is the verdict (green trust it, amber verify, red don't use). Range is the 95% interval for % right, counting each question once; n is questions × runs for the tools track.
- How it was run
- Claude Code on a Claude Pro subscription
- Tool loop
- Claude Code's own tool loop
- Runs
- 1 on its own, 3 with tools; rates pool all runs, ranges count each question once
- Verdict rule
- Trust it: wrong ≤ 5% and right ≥ 80%. Don't use it: wrong > 20%. Otherwise verify its answers.
Limits, harness differences and corrections: methodology.