tdg-personal/deep-research
Multi-source deep research using firecrawl and exa MCPs. Searches the web, synthesizes findings, and delivers cited reports with source attribution. Use when the user wants thorough research on any topic with evidence and citations.
76
76%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Advisory
Suggest reviewing before use
Quality
Discovery
77%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 solid description that clearly communicates both what the skill does and when to use it, with specific tool names and concrete actions. Its main weaknesses are moderate trigger term coverage (missing common user phrasings like 'look up' or 'find out about') and some conflict risk due to the broad scope of 'research on any topic'. The description uses proper third-person voice and avoids vague fluff.
Suggestions
Expand trigger terms to include natural user phrasings like 'look up', 'find information about', 'investigate', 'literature review', 'fact-check', or 'in-depth analysis'.
Add specificity to the 'when' clause to reduce conflict risk, e.g., 'Use when the user needs multi-source investigation, literature reviews, or comprehensive reports with citations—not for quick factual lookups.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists multiple concrete actions: searches the web, synthesizes findings, delivers cited reports with source attribution. Also names specific tools (firecrawl and exa MCPs). | 3 / 3 |
Completeness | Clearly answers both what ('searches the web, synthesizes findings, delivers cited reports with source attribution') and when ('Use when the user wants thorough research on any topic with evidence and citations'). | 3 / 3 |
Trigger Term Quality | Includes good terms like 'research', 'citations', 'evidence', and 'web' that users might naturally say. However, it misses common variations like 'look up', 'find information', 'sources', 'investigate', 'literature review', or 'fact-check' that users might also use. | 2 / 3 |
Distinctiveness Conflict Risk | The mention of specific MCP tools (firecrawl and exa) and the focus on 'deep research' with citations helps distinguish it, but 'research' is broad enough that it could overlap with simpler search or Q&A skills. The phrase 'any topic' further reduces distinctiveness. | 2 / 3 |
Total | 10 / 12 Passed |
Implementation
64%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a solid, actionable research skill with concrete MCP tool calls and a well-defined report template. Its main weaknesses are the lack of validation checkpoints within the workflow (e.g., what to do when searches return poor results) and some verbosity in sections like activation triggers and examples that Claude could infer. The content would benefit from integrating quality checks as explicit workflow gates rather than a separate rules section.
Suggestions
Add explicit validation checkpoints within the workflow, e.g., after Step 3 check if sufficient unique sources were found and retry with different queries if not, and after Step 4 verify source quality before proceeding to synthesis.
Trim or remove the 'When to Activate' and 'Examples' sections — these are trigger phrases and use cases that Claude can infer from the skill description and workflow.
Consider splitting the report template and quality rules into a referenced file (e.g., REPORT_TEMPLATE.md) to keep the main skill leaner and improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is mostly efficient but has some unnecessary padding. The 'When to Activate' section lists trigger phrases Claude could infer, the 'Examples' section at the end adds little value beyond what's already clear, and some explanatory text like 'Both together give the best coverage' is filler. However, the core workflow and code examples are reasonably tight. | 2 / 3 |
Actionability | Provides specific MCP tool calls with exact function signatures and parameters, a concrete report template with markdown structure, specific search strategies (2-3 keyword variations, 15-30 sources), and clear instructions for parallel execution with subagents. The guidance is directly executable. | 3 / 3 |
Workflow Clarity | The 6-step workflow is clearly sequenced and logical, but lacks explicit validation checkpoints. There's no step to verify source quality before synthesis, no feedback loop if searches return poor results, and no checkpoint between reading sources and writing the report. The quality rules are listed but not integrated as validation gates within the workflow. | 2 / 3 |
Progressive Disclosure | The content is well-structured with clear headers and sections, but it's a fairly long monolithic document (~120 lines of content). The report template, quality rules, and parallel research guidance could be split into referenced files. There are no references to external files for advanced usage or detailed examples. | 2 / 3 |
Total | 9 / 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 | |
Reviewed
Table of Contents