01Safety
30.0 / 30
What changed in the harness
Selection accuracy 100%, destructive-action safety rate 100% (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
29.2 / 30
03Economics
20.0 / 20
04Discoverability
13.6 / 20
Highest-impact fix
Estimated gain +6 pointsMake target tools discoverable on the first call
Clarify tool names, decision boundaries, and required argument schemas so an agent can choose and construct the target call without exploratory steps.
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.
4 pairs where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
compare_models |
recommend_model |
medium | A natural task like 'which of these two models should I use for coding under my budget' names candidate models (fits compare_models) but asks for a choice driven by use case/budget (fits recommend_model), so an agent could plausibly select either tool. |
list_top_models |
recommend_model |
medium | A bare query like 'what are the best models for coding or vision?' satisfies both list_top_models(category=...) returning a ranked table and recommend_model(use_case=...) returning picks with reasoning, and neither description forces a disambiguating constraint. |
get_model_info |
compare_models |
low | A task such as 'tell me about claude-sonnet-4.6 and gpt-5.1, including benchmarks and pricing' could be answered either by fetching per-model info or via a side-by-side comparison; shared capability/pricing vocabulary blurs the boundary. |
list_top_models |
compare_models |
low | 'Compare the top coding models' involves both ranking and comparison without naming models, so the agent might pick list_top_models for a ranked table or call compare_models with guessed model IDs. |
Compare the field