01Safety
0.0 / 30
What changed in the harness
Selection accuracy 91→91, 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
26.5 / 30
03Economics
19.8 / 20
04Discoverability
10.6 / 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.
5 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 |
|---|---|---|
aethyn_click |
no_return_description |
Click an element on the page, identified by a 'ref' from aethyn_snapshot (e.g. 'e12') or a fallback 'selector' (CSS, or Playwright role=/text= engine). Returns a status result indicating whether the click succeeded, with an error message if it did not. |
aethyn_type |
no_return_description |
Type text into an input field, identified by a 'ref' from aethyn_snapshot or a fallback 'selector'. Set submit:true to press Enter after typing. Returns a status result indicating whether the text was entered (and submitted, if requested) successfully. |
aethyn_new_identity |
no_return_description |
Rotate the session to a fresh residential exit IP (same country) and clear cookies; use between independent tasks or after a soft-block. Returns the session's new exit IP and confirmation that the identity rotation completed. |
aethyn_close |
no_return_description |
Close the session's browser context and free memory; call this when a task is done. Returns a confirmation that the session was closed. |
aethyn_list_countries |
no_return_description |
List available exit countries (and Elite-tier cities where known) so you can discover geo options at runtime before launching a session. Returns the list of available country codes, plus associated city names when with_cities is set or cities are known. |
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 |
|---|---|---|---|
aethyn_navigate |
aethyn_click |
medium | A task like 'go to the About page' or 'open the login screen' could be satisfied either by navigating directly to a known URL or by clicking a link/button with that name; an agent might pick click when navigate is more appropriate (or vice versa) since both can accomplish page transitions. |
aethyn_launch_browser |
aethyn_new_identity |
medium | Both tools are about obtaining a fresh proxy/exit identity (shared tokens 'country','exit','residential'); a task like 'get me a new IP' or 'start fresh in Germany' is ambiguous between launching an entirely new session (launch_browser) and rotating the exit IP within an existing session (new_identity), especially if the agent isn't certain whether a session already exists. |
Compare the field