mgrep-code-search
Semantic code search using mgrep for efficient codebase exploration. This skill should be used when searching or exploring codebases with more than 30 non-gitignored files and/or nested directory structures. It provides natural language semantic search that complements traditional grep/ripgrep for finding features, understanding intent, and exploring unfamiliar code.
84
77%
Does it follow best practices?
Impact
93%
11.62xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./skills/mgrep-code-search/SKILL.mdQuality
Discovery
89%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 well-crafted description that clearly communicates both what the skill does and when to use it, with a specific threshold for activation (30+ files). It includes good trigger terms and is distinctive from traditional search tools. The main weakness is that the specific capabilities could be more concretely enumerated rather than described in somewhat abstract terms like 'understanding intent.'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (semantic code search) and the tool (mgrep), and mentions some actions like 'finding features, understanding intent, and exploring unfamiliar code,' but these are somewhat abstract rather than listing multiple concrete discrete actions. | 2 / 3 |
Completeness | Clearly answers both 'what' (semantic code search using mgrep for codebase exploration, natural language semantic search complementing grep/ripgrep) and 'when' (when searching codebases with more than 30 non-gitignored files and/or nested directory structures). The 'when' clause is explicit and specific. | 3 / 3 |
Trigger Term Quality | Includes strong natural trigger terms: 'search', 'exploring codebases', 'semantic search', 'grep', 'ripgrep', 'code search', 'unfamiliar code', 'nested directory structures'. These cover a good range of terms users would naturally use when needing this skill. | 3 / 3 |
Distinctiveness Conflict Risk | Clearly distinguishable from generic search or grep skills by specifying semantic/natural language search via mgrep, with a specific threshold (30+ files, nested directories) that differentiates it from simpler search approaches. Unlikely to conflict with traditional grep/ripgrep skills. | 3 / 3 |
Total | 11 / 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 skill with clear executable examples and good command documentation. Its main weaknesses are moderate verbosity (the 'When to Use' section and repeated notes) and a workflow that lacks error handling or validation steps. The content is well-structured but could be tightened to better respect token budget.
Suggestions
Remove or significantly trim the 'When to Use This Skill' section — this decision context is already provided in the skill's metadata/description and Claude can infer appropriate usage.
Add a brief validation step to the workflow, e.g., how to verify the watcher/index is running and current before searching (check for errors or stale index).
Consolidate the 'Important Notes' section into the relevant workflow steps to avoid repetition (e.g., the 'always use bunx' note and 'run watch first' are already stated above).
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Generally efficient but includes some unnecessary sections like the 'When to Use This Skill' guidance (Claude can infer this from context) and the environment variables table which adds bulk without high value. The 'Important Notes' section repeats information already stated (e.g., running watch before searching). | 2 / 3 |
Actionability | Provides fully executable, copy-paste ready commands throughout. The search examples are concrete with real natural language queries, options are clearly documented with a table, and the invocation pattern is explicit and consistent. | 3 / 3 |
Workflow Clarity | The workflow section provides a clear 3-step sequence (start watcher → search → refine), but lacks validation checkpoints. There's no guidance on what to do if indexing fails, if the watcher isn't running, or how to verify the index is current before searching. For a non-destructive search tool this is less critical, but the skill could benefit from error handling guidance. | 2 / 3 |
Progressive Disclosure | Content is reasonably well-structured with clear sections and headers, but everything is inline in a single file. The environment variables and detailed options tables could be referenced separately. For a skill of this length (~90 lines), the organization is adequate but the 'When to Use' section and env vars section inflate the main file unnecessarily. | 2 / 3 |
Total | 9 / 12 Passed |
Validation
100%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
- Repository
- intellectronica/agent-skills
- Commit
9b0e00a
- 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.