semantic-scan
Build a computation register and detect semantic duplicates across architectural layers. Finds business logic reimplemented multiple times in different layers — the same domain calculation independently appearing in domain services, client adapters, and presentation components. Runs incrementally (git-diff-based) after the first scan. Produces a structured duplicate report with file:line references and canonical location suggestions.
Source facts
- Repository
- bdfinst/agentic-dev-team
- Last source activity
- July 31, 2026 at 12:40
- Detected SKILL.md language
- English
- Stars
- 299
- Forks
- 36
Install options
The review-first prompt is selected by default. You can switch to a direct command or download a local copy.
Review the source files
Read SKILL.md and any companion files shown by SkillsMP before deciding whether to install.
Showing SKILL.md
SKILL.md
Source instructions · Read-only preview- name
- semantic-scan
- description
- Build a computation register and detect semantic duplicates across architectural layers. Finds business logic reimplemented multiple times in different layers — the same domain calculation independently appearing in domain services, client adapters, and presentation components. Runs incrementally (git-diff-based) after the first scan. Produces a structured duplicate report with file:line references and canonical location suggestions.
- argument-hint
- [path] [--full] [--no-opus]
- user-invocable
- true
# Semantic Scan
Role: worker.
## Worker constraints
1. Detect and report duplicates; do not refactor.
2. Run incrementally (git-diff based) after the first scan.
3. **Be concise.** Report the duplicate table with file:line refs, no narration.
## Steps
### 1. Parse arguments
Capture the optional path and flags from `$ARGUMENTS`.
### 2. Delegate
Invoke the semantic-duplication-scan skill.
**Graph-assisted discovery.** Prefer CodeGraph/Repowise over raw `Grep`/`Read`
when locating and comparing candidate implementations — Repowise's semantic
search is especially well suited to finding duplicate implementations across
layers. See
[`knowledge/codegraph-vs-graphify.md`](../../knowledge/codegraph-vs-graphify.md)
for tool selection and the fallback contract.
### 3. Report
Output the duplicate report.
Apply the guidelines defined in skills/semantic-duplication-scan/SKILL.md to the current task. Read the skill file and follow its process flow, pre-filter rules, annotation procedure, clustering strategy, and report format.
## Flags
- `[path]` — Optional subdirectory to scope the scan. Only files under this path are re-annotated; out-of-scope register entries are preserved.
- `--full` — Force full-scan mode regardless of whether a register exists. Bypasses the shallow-clone pre-flight check. Use when `lastScanCommit` is stale or after a major restructuring.
- `--no-opus` — Skip Opus canonical resolution for ambiguous clusters. Ambiguous clusters are reported as `canonical: ambiguous — human review required` instead of triggering an Opus call. Use in cost-sensitive environments (CI, personal accounts).
Apply this skill to: $ARGUMENTS
View on GitHub