Loading ToolBench
Deal-mechanics interpretation for brand deals, UGC, and newsletter sponsorships: pricing models, usage rights, whitelisting, exclusivity, negotiation moves, sponsor vetting. Structure, not rate quotes.
Creator Deals MCP has 6 well-intentioned tools with clear domain focus (deal mechanics for creators). All tools have descriptions (100-250 chars, within baseline 194 avg) and input schemas with type definitions. However, critical gaps reduce production readiness: (1) Output schemas are completely undocumented, no tool declares what it returns, forcing LLMs to guess structure; (2) Parameter descriptions lack actionable constraints (enums, ranges, formats), e.g., 'model_type' accepts 5 values but no enum constraint; (3) No error handling guidance, tools don't explain failure modes or recovery paths; (4) Tool names lack action verbs (interpret_*, analyze_*, scope_* are descriptive but not imperative); (5) No pagination, limits, or result-shaping for tools that could return large outputs (e.g., scope_checklist, deal_structure_template). Naming is clear and single-responsibility is respected, but schema completeness and error guidance are missing.
Analyzes usage rights clauses (organic only, paid usage, whitelisting, exclusivity, raw footage, payment terms) and returns: what each term means, its relative value as a multiplier of base rate, and common traps.
Returns a deal-structure template for a given deal type (brand deal, UGC, newsletter sponsorship, affiliate/CPA, retainer). Includes sections for deliverables, usage, exclusivity, payment, timeline, and approval. Use as a starting point for your own contracts.
Interprets a pricing model (flat fee, CPM, performance-based, per-click, retainer) and returns structural analysis: what it means, when it favors you vs. the brand, key rules, and counter-patterns.
Given a negotiation situation (lowball offer, budget is fixed, scope creep, payment delay, exclusivity overreach, etc.), returns the recommended move, what to say, and what NOT to do.
Returns a structured checklist of 10 critical deal-scoping questions: deliverables, revisions, usage, exclusivity, approval, timeline, payment, raw assets, reporting, and disclosure. Use this to audit a brief or offer before responding.
Evaluates a sponsor/brand based on red flags and green flags: audience alignment, payment history, contract clarity, usage rights respect, and exclusivity scope. Returns a risk assessment and recommendations.
Output schemas completely undocumented. No tool declares what fields it returns, forcing LLMs to infer structure from examples or fail silently.
Input parameters lack enum constraints and format specifications. 'model_type' accepts 5 known values but no enum; 'usage_type' and 'deal_type' similarly unconstrained. LLMs will hallucinate invalid values.
No error handling guidance. Tools do not document failure modes (invalid model_type, malformed clause_text, etc.) or recovery paths. LLMs cannot self-correct on invalid input.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 40 | 2026-07-28+ | v2 |
Tool names use descriptive verbs (interpret_, analyze_, scope_) rather than action verbs (get_, create_, describe_). Reduces clarity for LLM tool selection and does not follow baseline naming convention (90% of A+ tools start with get/list/create/search/update).
sponsor_vetting accepts red_flags and green_flags as free-form string arrays with no validation. No guidance on expected format, length, or content structure. LLMs may pass malformed or irrelevant flags.