research-curator
Manage research entries in ./research/ — create, refresh, and validate. Use when asked to add a tool, "document this", "research this", "refresh this research", "validate research entries", or given a tool URL. Modes: default (single URL), --batch (multiple URLs in parallel), --rerun (refresh stale entries), --validate (structural check and auto-fix).
87
85%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly communicates what the skill does (manage research entries with create/refresh/validate operations), when to use it (with explicit trigger phrases), and how it operates (four distinct modes). The description is concise yet comprehensive, with natural trigger terms and a clear niche that minimizes conflict risk.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple specific concrete actions: create, refresh, validate research entries, plus describes four distinct modes (default single URL, batch multiple URLs, rerun stale entries, validate with structural check and auto-fix). | 3 / 3 |
Completeness | Clearly answers both 'what' (manage research entries — create, refresh, validate in ./research/) and 'when' (explicit 'Use when...' clause with multiple trigger scenarios). Also describes operational modes for additional clarity. | 3 / 3 |
Trigger Term Quality | Includes natural trigger phrases users would actually say: 'add a tool', 'document this', 'research this', 'refresh this research', 'validate research entries', 'tool URL'. These cover a good range of natural language variations. | 3 / 3 |
Distinctiveness Conflict Risk | Highly specific niche: managing research entries in a specific directory (./research/) with distinct modes and triggers tied to tool URLs and research documentation. Unlikely to conflict with generic document or code skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
70%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a highly actionable and well-structured orchestration skill with excellent workflow clarity through Mermaid diagrams and explicit validation checkpoints. Progressive disclosure is handled well with appropriate references to external files. However, the skill is significantly over-engineered in verbosity — the agent relay rules with example tables, repeated pre-relay checklist references, dual representation of workflows (Mermaid + prose), and extensive output templates consume far more tokens than necessary for Claude's capabilities.
Suggestions
Consolidate the Agent Result Relay Rules into a compact bullet list (e.g., 'Preserve exact counts, failure reasons, file paths, and structured sections verbatim — never generalize or paraphrase') instead of the verbose table format with positive/negative examples.
Remove the prose re-explanation of workflows that are already fully captured in Mermaid diagrams (e.g., the 'Single Entry Rerun' numbered steps duplicate the Mermaid flowchart above them).
Reduce the pre-relay quality checklist to a single reference rather than repeating it in every mode section — define it once and reference by name.
Trim output format templates to show one complete example and note that other modes follow the same pattern with mode-specific fields, rather than spelling out all four templates in full.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~350+ lines. It over-explains relay rules with extensive tables of examples Claude could infer, repeats the pre-relay checklist concept many times, duplicates workflow steps in both Mermaid diagrams and prose, and includes lengthy output format templates. Significant token savings are possible without losing clarity. | 1 / 3 |
Actionability | The skill provides fully concrete, executable guidance: exact bash commands for validation/lint/commit/push, precise Agent tool invocation parameters with prompt strings, specific file paths, and detailed output format templates. Every mode has copy-paste ready instructions. | 3 / 3 |
Workflow Clarity | Multi-step workflows are exceptionally well-sequenced with authoritative Mermaid flowcharts, explicit decision points, validation checkpoints (pre-relay quality checklist), error handling (stop on failure, relay exact reasons), and feedback loops (validate → fix → re-validate pattern). Wave spawning has clear concurrency limits. | 3 / 3 |
Progressive Disclosure | Content is well-structured with clear one-level-deep references to external files (entry-template.md, validation-rules.md, batch-mode.md, agent files). The main skill provides summaries of each mode while pointing to detailed references. Navigation is easy with a dedicated Reference Links section. | 3 / 3 |
Total | 10 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- Jamie-BitFlight/claude_skills
- Commit
11ec483
- Reviewed
Table of Contents
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.