01Safety
0.0 / 30
What changed in the harness
Selection accuracy 100→100, token cost up 4%, 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
28.8 / 30
03Economics
20.0 / 20
04Discoverability
12.5 / 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.
2 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 |
|---|---|---|
vibie_create_site |
no_return_description |
Create a NEW Vibie site by uploading a local folder. Use this when the user wants to publish a brand new site, or when no .vibie/site.json marker exists in the folder. Vibie hosts static HTML/CSS/JS at a permanent URL like my-site-a3f9.vibie.page. Auto-creates a .vibie/site.json marker in the folder for future updates. Returns the newly created site's slug and permanent URL. |
vibie_update_site |
no_return_description |
Update an existing Vibie site with new content from a local folder. Use this when the user wants to re-deploy or push changes to an already-published site. If .vibie/site.json marker exists in the folder, the slug is read automatically; otherwise pass slug explicitly. Same URL, new content. Returns the updated site's slug and URL. |
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 |
|---|---|---|---|
vibie_create_site |
vibie_update_site |
high | A task like 'deploy/publish/host my site' is genuinely ambiguous because the agent must determine whether the folder is new (create) or already published (update). The descriptions route on the .vibie/site.json marker, but a user rarely states that, so the agent can easily pick the wrong one. |
vibie_list_sites |
vibie_get_site |
medium | A singular-request task like 'show me info about my site' or 'what's my site's URL' could lead to choosing get_site (which requires a slug the user may not have provided) versus list_sites (which discovers all sites). Without the slug being explicit, the choice is genuinely ambiguous. |
Compare the field