claims

Claims-based authorization for agents and operations. Grant, revoke, and verify permissions for secure multi-agent coordination. Use when: permission management, access control, secure operations, authorization checks. Skip when: open access, no security requirements, single-agent local work.

1.79x

Quality

84%

Does it follow best practices?

Impact

97%

1.79x

Average score across 3 eval scenarios

Securityby

Advisory

Suggest reviewing before use

Evaluation results

93%

44%

Multi-Agent Pipeline Permission Setup

Claim grant with scoped permissions

Criteria

Without context

With context

Grant command syntax

70%

100%

--agent flag used

100%

--claim flag used

100%

--scope flag used

70%

80%

Read claim for ingest

50%

100%

Read+write for transform

50%

100%

Spawn claim for orchestrator

50%

100%

Memory claim scoped to namespace

20%

100%

Path scopes not wildcarded

20%

100%

No over-permissioning

100%

50%

Correct claim types only

10%

100%

53%

Secure Deployment Agent Workflow

Temporary elevated access lifecycle

Criteria

Without context

With context

Grant command present

100%

Check command present

100%

Revoke command present

100%

Revoke after work

100%

Write claim granted

50%

100%

Execute claim granted

50%

100%

Scope to config path

60%

100%

Elevated claims revoked

70%

100%

Baseline read retained

100%

Correct --agent flag

100%

Revoke rationale documented

83%

100%

98%

30%

Agent Permission Audit and Cleanup

Claim audit and revocation

Criteria

Without context

With context

List command referenced

100%

Revoke command syntax

100%

Admin claim revoked

70%

100%

Excessive write revoked from reporter

62%

100%

Overly broad read scope fixed for reporter

75%

100%

Spawn revoked from notifier

62%

100%

Memory revoked from notifier

62%

100%

Correct --agent flag per agent

100%

No over-revocation

100%

Audit report identifies excesses

100%

Least privilege reasoning

100%

83%

Repository: ruvnet/ruflo
Commit: 398f7c2

Evaluated: about 1 month ago
Agent: Claude Code
Model: Claude Sonnet 4.6

Table of Contents

Multi-Agent Pipeline Permission Setup Secure Deployment Agent Workflow Agent Permission Audit and Cleanup

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.