semantic-szz-analyzer
Extends classic SZZ with semantic code understanding to reduce false positives and improve accuracy of bug-introducing commit identification. Use after classic SZZ has produced candidates, when SZZ precision is too low for the task, or when the user needs high-confidence bug-introduction data.
Install with Tessl CLI
npx tessl i github:santosomar/general-secure-coding-agent-skills --skill semantic-szz-analyzer85
Quality
81%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Discovery
85%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-structured skill description with clear capabilities and explicit trigger conditions. The main weakness is the reliance on technical jargon (SZZ) that users may not naturally use when seeking this functionality. The description would benefit from including more accessible trigger terms alongside the technical ones.
Suggestions
Add natural language trigger terms users might say, such as 'find which commit introduced a bug', 'git blame analysis', 'trace bug origin', or 'identify faulty commits'
Consider briefly explaining what SZZ is for users unfamiliar with the algorithm, or add alternative phrasings like 'bug origin detection' or 'commit fault localization'
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists specific concrete actions: 'Extends classic SZZ with semantic code understanding', 'reduce false positives', 'improve accuracy of bug-introducing commit identification'. These are clear, actionable capabilities. | 3 / 3 |
Completeness | Clearly answers both what ('Extends classic SZZ with semantic code understanding to reduce false positives') AND when ('Use after classic SZZ has produced candidates, when SZZ precision is too low, or when user needs high-confidence bug-introduction data'). | 3 / 3 |
Trigger Term Quality | Includes domain-specific terms like 'SZZ', 'bug-introducing commit', 'false positives', but these are technical jargon. Missing more natural user terms like 'git blame', 'find bug source', 'trace bug origin', or 'commit analysis'. | 2 / 3 |
Distinctiveness Conflict Risk | Very specific niche targeting SZZ algorithm enhancement. The explicit mention of 'classic SZZ', 'bug-introducing commit identification', and the specific use cases make it highly distinguishable from other skills. | 3 / 3 |
Total | 11 / 12 Passed |
Implementation
77%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is a well-structured skill that clearly explains the semantic enhancement over classic SZZ with excellent workflow clarity and token efficiency. The main weakness is the lack of executable code—the filters and re-ranking are described conceptually but Claude would need to implement them from scratch. The worked example partially compensates by showing the reasoning process.
Suggestions
Add executable code snippets for at least one filter (e.g., AST-diff using tree-sitter or similar) to make the guidance more actionable
Provide a concrete command or script invocation showing how to run semantic SZZ after classic SZZ completes
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every section earns its place—explains the delta over classic SZZ without rehashing what Claude knows about git or blame. The tables and examples are dense with information, no padding or unnecessary context. | 3 / 3 |
Actionability | Provides clear conceptual guidance with a worked example and output format, but lacks executable code. The filters are described abstractly (e.g., 'AST-diff meaningfulness') without concrete implementation commands or scripts to run. | 2 / 3 |
Workflow Clarity | Clear sequence: run classic SZZ first → apply semantic filters → re-rank survivors. The worked example demonstrates the full workflow with explicit decision points. The 'Do not' section provides important guardrails. | 3 / 3 |
Progressive Disclosure | References the parent skill `szz-bug-identifier` appropriately, but all content is inline. The filter table, re-ranking criteria, worked example, and edge cases could benefit from being split into separate reference files for a skill of this complexity. | 2 / 3 |
Total | 10 / 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.
- 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.