Public leaderboard

Public assessment

optimaquantum/claude-critical-rules-mcp (@optima-quantum/claude-critical-rules-mcp-new)

optima-quantum-claude-critical-rules-mcp-new · v1.2.0 · scanned

What changed in the harness

Selection accuracy 100→100, token cost up 18%, 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.

01Safety

0.0 / 30

0.0 out of 30
02Legibility

28.2 / 30

28.2 out of 30
03Economics

20.0 / 20

20.0 out of 20
04Discoverability

12.5 / 20

12.5 out of 20

Highest-impact fix

Estimated gain +30 points

Add 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
verify_compliance
no_return_description
Verify that all mandatory checks required by the critical rules have been completed before starting a technical task, and return a report indicating whether the task is approved to start, listing any required checks that are still missing or have failed so they can be addressed first.
get_rules_summary
name_restates_behavior no_return_description
Return a concise overview of the critical rules grouped by category, with the main requirement of each category, so you can quickly review which rules apply before and during a technical task; use this instead of retrieving the full rule text when you only need the key requirements.
get_version_info
no_return_description
Return the currently installed version of the critical rules along with the result of a check for newer versions, including the latest version number so you can decide whether an update is warranted.
check_for_updates
no_return_description
Check GitHub for a newer release of the critical rules and return whether an update is available, including the latest version number, so you know whether to run update_rules.
update_rules
no_return_description
Download and install the latest version of the critical rules from GitHub and return the outcome of the update, including the newly installed version or a note that the rules are already up to date, so you can confirm the update succeeded.

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
check_for_updates update_rules medium Both tools operate identically by checking GitHub for the latest critical-rules version, so a task like 'make sure my rules are up to date' or 'see if I need to update' could resolve to either. An agent may call the read-only checker when the user wants installation, or vice versa, since the descriptions overlap in purpose and the action boundary (check vs. install) is the only distinction.

Compare the field

One score is useful.
The evidence makes it actionable.

Back to the leaderboard