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, company & market intelligence, baseline knowledge.
| Job | Closed-book | With Orbit Sentinel tools | n | ||||
|---|---|---|---|---|---|---|---|
| right / wrong | range | declined | right / wrong | range | declined | ||
| Catalog lookup | 58 / 24 | 43–71 | 18 | 98 / 2 | 88–100 | 0 | 45×3 |
| Constellation & fleet status | 49 / 12 | 35–63 | 40 | 98 / 2 | 88–100 | 1 | 43×3 |
| Debris & conjunction context | 45 / 17 | 31–60 | 38 | 93 / 4 | 81–98 | 3 | 42×3 |
| Regulatory & licensing | 28 / 10 | 16–43 | 63 | 47 / 14 | 32–62 | 39 | 40×3 |
| Company & market intelligence | 41 / 27 | 28–57 | 32 | 83 / 2 | 69–91 | 15 | 41×3 |
| Ground infrastructure | 18 / 27 | 7–39 | 55 | 74 / 17 | 53–88 | 9 | 22×3 |
| Baseline knowledge | 100 / 0 | 87–100 | 0 | 100 / 0 | 87–100 | 0 | 25×3 |
| Orbit Sentinel-only questions | 0 / 8 | 0–23 | 92 | 77 / 10 | 50–92 | 13 | 13×3 |
% 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.