Loading ToolBench
Content Repurposing MCP for creators. One asset, every platform: repurpose maps, platform specs, atomic-content extraction, and the film-once-cut-many workflow. Never repost the identical file again.
6 tools with clear naming (verb_noun pattern: get_*, extract_*) and reasonable descriptions (100-200 chars). However, ALL tools have empty input schemas (no properties defined beyond the parameter itself), violating the critical schema requirement. Parameters lack type definitions and descriptions. No output schemas documented. Error handling absent. Tools are read-only reference data, not stateful operations, which limits complexity but does not excuse schema gaps.
Given a target platform, returns the adaptation rules to reshape a piece of content for it (aspect, length, hook approach, caption style, hashtags, what NOT to do). Platforms: tiktok, reels, shorts, reel_carousel, x_thread, linkedin, newsletter.
The 6 atomic content types you can pull from one long asset (pull quotes, clips, stats/receipts, steps/lists, story beats, questions) and what each becomes. Use to break one asset into many posts.
The complete repurposing library in one payload: repurpose maps, platform specs, atomic types. For full agent context.
Platform specs for adapting content: aspect ratio, ideal length, caption style, hashtag norms, hook rule, cadence note. Platforms: tiktok, reels, shorts, reel_carousel, x_thread, linkedin, newsletter. Omit for all.
Given a source content type (long_video, podcast_episode, blog_or_essay, single_short_video), returns every format it can become and exactly how to adapt it. Omit source for all maps.
All 6 tools have empty or missing input schemas. Parameters declared but no type, description, or constraint metadata. Violates critical schema requirement: 'If a tool has NO input schema at all: its schema score MUST be 0.'
No output schemas documented for any tool. LLMs cannot plan downstream calls or extract structured data. Violates pattern: 'Document the output schema. LLMs need to know what fields to expect.'
No error handling or recovery guidance. Tools return static data, but no documentation of failure modes, validation errors, or what to do if a platform is unsupported.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 43 | 2026-07-28+ | v2 |
The 'film once, cut many' batch workflow: capture one strong asset, extract atomic pieces, adapt each per platform, schedule across the week. Returns the step-by-step.
Parameter descriptions missing or trivial. 'source' and 'platform' params have minimal context. No guidance on valid values, format, or when to use each tool vs. alternatives.
No tool annotations (readOnlyHint, destructiveHint, idempotentHint). All tools are read-only reference data, should declare readOnlyHint: true to signal safety to agents.