TypeScript prompt library for Claude Code — battle-tested prompts, CLI, MCP, and test harness
Two tools with basic schemas and descriptions, but significant gaps in parameter documentation and output schema clarity. list_prompts has no parameters documented; run_prompt has three parameters but only two are described (dryRun lacks description). Neither tool documents return structure. Descriptions are present but generic (10 - 20 words each). No error handling guidance, no examples of valid input/output, no pagination support despite potential for large prompt lists. Tool names follow verb_noun convention (list_, run_) but descriptions do not explain when to use each tool or what data structures are returned.
List all prompt ids, categories, and short descriptions.
Run a prompt by id with a JSON object matching that prompt's input schema (see project README).
Output schemas not documented. Neither tool specifies what fields are returned or their types. LLMs cannot plan downstream calls or extract data reliably.
Parameter descriptions incomplete. run_prompt's 'dryRun' parameter has no description. 'input' parameter description is vague ('JSON payload for the prompt'), does not specify required fields, format, or constraints.
list_prompts has empty input schema but no description of what it returns. Does it return prompt IDs, full definitions, or summaries? Pagination not mentioned despite potential for large lists.
No error handling guidance. If run_prompt fails (invalid prompt ID, malformed input, API error), LLM receives no recovery hints. No distinction between retryable and fatal errors.
| Scored | Grade | Overall | Spec posture | Rubric |
|---|---|---|---|---|
| 2026-09-23 | F | 42 | 2025-06-18+ | v2 |
Tool descriptions lack context for selection. 'Run a prompt by id' does not explain when to call this vs list_prompts, what prerequisites exist, or what the output is used for.