01Safety
0.0 / 30
What changed in the harness
Selection accuracy 98%, destructive-action safety rate 0% (baseline only -- no rewrite pass applied).
Category breakdown
Where the score comes from.
Earned points across the four signals Gradable measures. Safety and Legibility are scored out of 30; Economics and Discoverability are scored out of 20.
02Legibility
29.1 / 30
03Economics
18.5 / 20
04Discoverability
15.9 / 20
Highest-impact fix
Estimated gain +30 pointsAdd explicit identity and permission preflight tools
Expose machine-readable principal/tenant confirmation and a non-mutating permission check so agents can verify both before destructive actions.
Description evidence
Defects and rewrites.
0 defects found across the exposed tool descriptions. Suggested rewrites make purpose, inputs, boundaries, and returns easier for an agent to understand.
| Tool | Defect types | Suggested rewrite |
|---|---|---|
| No description defects were flagged in this assessment. | ||
Selection evidence
Confusable tool pairs.
2 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
prior_contribute |
prior_feedback |
medium | Both capture information going back into Prior, so a task like 'give feedback on this fix' is ambiguous: it could mean rating a prior search result (prior_feedback) or sharing the solution itself (prior_contribute), especially since the contribute prompt is phrased as an inline feedback-style question ('want me to contribute this?'). |
prior_search |
prior_feedback |
medium | The search tool's own description and schema expose a previousSearchFeedback parameter for rating the last search, so a task like 'search for this error and rate the result' could wrongly route rating params through prior_search instead of the dedicated prior_feedback tool, or call prior_feedback expecting it to perform a search with feedback. |
Compare the field