01Safety
0.0 / 30
What changed in the harness
Selection accuracy 98→100, token cost up 1%, unconfirmed writes 0%→0%.
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
29.0 / 30
03Economics
16.5 / 20
04Discoverability
18.6 / 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.
1 defect 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 |
|---|---|---|
get_meme_template |
no_return_description |
Fetch one meme template by its slug (from search_meme_templates or a memebo.at/templates/<slug> URL). Returns the template's dimensions, image URL, suggested caption count, and its list of available background styles (availableStyles) that can be passed to create_meme's style parameter. |
Selection evidence
Confusable tool pairs.
1 pair where similar names or overlapping descriptions may send an agent toward the wrong tool.
| Tool A | Tool B | Confidence | Why they collide |
|---|---|---|---|
search_meme_templates |
get_meme_template |
medium | A request like "find the drake meme template and show its details" could lead an agent to call either the search tool (which already returns image URL and caption count) or fetch a specific template by slug, especially if the agent guesses a slug instead of searching first. |
Compare the field