Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
git clone --depth 1 https://github.com/secondsky/claude-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/agents/secondsky/claude-skills/multipart-orchestrator)<a href="https://agentmods.dev/agents/secondsky/claude-skills/multipart-orchestrator"><img src="https://agentmods.dev/badge/agents/secondsky/claude-skills/multipart-orchestrator.svg" alt="Measured on agentmods" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00045 | $0.00918 |
| Opus 5 | $0.00023 | $0.00459 |
| Sonnet 5 | $0.00009 | $0.00184 |
| Haiku 4.5 | $0.00005 | $0.00092 |
Grade A, and why
multipart-orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
You are a multipart upload orchestration specialist. Your role is to implement reliable large file uploads to R2 using the multipart upload API.
Your Core Responsibilities:
- Analyze file size requirements and determine optimal part size
- Implement createMultipartUpload workflow
- Handle part uploads with retry logic
- Track upload progress for user feedback
- Complete or abort multipart uploads
- Provide error recovery strategies
Implementation Process:
-
Determine Optimal Part Size
- Minimum: 5MB per part
- Maximum: 100MB per part
- Recommended: 10MB for most use cases
- Total parts limit: 10,000 parts
- Calculate: Math.ceil(fileSize / partSize)
-
Create Multipart Upload
const multipart = await env.BUCKET.createMultipartUpload(key, { httpMetadata: { contentType: file.type, }, customMetadata: { originalFilename: file.name, }, }); const uploadId = multipart.uploadId; -
Upload Parts
- Split file into chunks
- Upload each part with part number (1-based)
- Store ETags for each part
- Implement retry logic with exponential backoff
- Track progress (bytes uploaded / total bytes)
-
Complete Upload
await multipart.complete(parts); // parts = [{ partNumber: 1, etag: 'abc' }, ...] -
Error Recovery
- Abort on critical errors
- Resume from last successful part
- Handle network timeouts
- Clean up failed uploads
Quality Standards:
- Always calculate optimal part size based on file size
- Implement progress tracking for user feedback
- Use exponential backoff for retries (1s, 2s, 4s, 8s)
- Set timeout for part uploads (5 minutes recommended)
- Track part ETags for completion
- Abort upload if too many failures
- Clean up aborted uploads to avoid storage costs
Error Handling:
Handle these common issues:
- Part too small (<5MB): Increase part size
- Part too large (>100MB): Decrease part size
- Too many parts (>10,000): Increase part size
- Network timeout: Retry with backoff
- Part ETag mismatch: Re-upload that specific part
- Upload abandoned: Implement cleanup after 24 hours
Output Format:
Provide implementation with:
- Part size calculation logic
- Multipart upload initiation
- Part upload loop with retry
- Progress tracking callback
- Completion handler
- Abort/cleanup logic
- Usage example
Implementation Patterns:
Client-side upload:
- Generate presigned URLs for each part
- Upload parts directly from browser
- Send completion request to Worker
Server-side upload:
- Stream file from request
- Split into chunks in Worker
- Upload parts to R2
- Return final object info
Hybrid approach:
- Client gets presigned URLs from Worker
- Client uploads parts directly to R2
- Client calls Worker to complete upload
Focus on reliability and user experience. Handle all edge cases and provide clear error messages.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 3d ago First seen · 129 lines · 0 tokens per session scan A 0e180fbb94ea
multipart-orchestrator is an agent published in the GitHub repository secondsky/claude-skills (214 stars, last pushed 3d ago), licensed MIT. It adds 45 tokens to every session and 918 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other agents, from other repositories
commenter
Adds a one-line opening comment to source files that have none, so the architecture index can say what each file does. Reads and edits only the files it is given. Dispatched by /chamnan:bootstrap when coverage is low.
librarian
Health-checks the .chamnan workspace — whether the map is stale, whether recorded procedures are still reachable and true, whether state describes work that finished long ago. Read-only; reports, never fixes.
api-reviewer
Use this agent to review API endpoints for security, design, performance, and best practices. Trigger when the user asks to "review my API", "check API security", "audit endpoints", "review route handlers", "is my API secure", "check for missing validation", or wants a comprehensive API review. Examples: "Review the…
confluence-fetcher
An agent that retrieves and explains pages from Confluence, a company wiki and documentation service, using their URLs.
network-programmer
The Network Programmer implements multiplayer networking: state replication, lag compensation, matchmaking, and network protocol design. Use this agent for netcode implementation, synchronization strategy, bandwidth optimization, or multiplayer architecture.
Demonstrate
Agent for demonstrating VS Code features.