feat(edgeparse): add EdgeParse plugin to marketplace#152
Conversation
Add EdgeParse as a built-in plugin providing Rust-native PDF extraction that converts PDFs to Markdown/JSON/HTML/text. Includes a skill for intelligent activation, marketplace registration, and updates MarkItDown skill to recommend EdgeParse for PDF use cases.
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
There was a problem hiding this comment.
No issues found across 9 files
Confidence score: 5/5
- Automated review surfaced no issues in the provided summaries.
- No files require special attention.
Architecture diagram
sequenceDiagram
participant User as User / CLI
participant Market as Marketplace (.claude-plugin/marketplace.json)
participant Agent as Claude Agent
participant MID as MarkItDown Plugin
participant EP as NEW: EdgeParse Plugin
participant FS as File System (PDF)
Note over User, Market: Plugin Discovery & Setup
User->>Market: /plugin install edgeparse
Market-->>User: NEW: Register EdgeParse as a "document" skill
Note over Agent, FS: PDF Extraction Flow
User->>Agent: "Extract tables from report.pdf"
Agent->>MID: Evaluate tool for PDF task
MID-->>Agent: CHANGED: Recommend EdgeParse for structured/complex PDFs
Agent->>EP: NEW: convert(path, format="markdown", table_method="cluster")
EP->>FS: Attempt to read PDF
alt PDF is password protected
FS-->>EP: Encrypted stream
EP-->>Agent: Error: Password required
Agent-->>User: Request PDF password
else PDF is accessible
FS-->>EP: PDF binary data
EP->>EP: NEW: Rust-native deterministic parsing
opt format="json"
EP->>EP: Generate bounding boxes & metadata
end
EP-->>Agent: Return structured content (MD/JSON/HTML)
Agent-->>User: Display parsed document content
end
Note over EP, FS: Batch Processing Pattern
opt Batch conversion
EP->>EP: NEW: Run concurrent extractions via async executor
end
There was a problem hiding this comment.
Code Review
This pull request introduces the edgeparse plugin for deterministic PDF extraction, including marketplace registration, comprehensive documentation, and integration patterns. The author field in plugin.json should be updated to credit the project organization rather than an individual contributor. Additionally, the inclusion of files within the .agents/skills/ directory violates repository policy for externally managed skills; since skills-lock.json is present, these files should be managed in the upstream repository and removed from this pull request.
Summary
.claude-plugin/marketplace.jsonandrelease-please-config.jsonREADME.mdTest plan
plugins/edgeparse/.claude-plugin/plugin.jsonis valid (claude plugin validate plugins/edgeparse).claude-plugin/marketplace.jsonSummary by cubic
Add
edgeparseas a built-in plugin for fast, deterministic PDF extraction to Markdown/JSON/HTML/text. Improves PDF workflows by surfacing the plugin in the marketplace/docs and guiding users to use it for PDFs.plugins/edgeparseskill with API and integration docs, plugin manifest, and skills lock.edgeparsein.claude-plugin/marketplace.jsonand set up releases inrelease-please-config.json.edgeparsefor PDF tasks (page ranges, borderless tables, structured output).Written for commit 50f8502. Summary will update on new commits.