Nicheloom

Market intelligence for builders — see what's gaining traction before it's crowded.

Context Mode

315 KB of MCP output becomes 5.4 KB in Claude Code

Details

External ID
47148025
Source
HN
Company
—
Product
Context Mode
Website domain
github.com
Launched
Feb. 25, 2026
Cohort
—
Upvotes
84
Upvotes percentile
0.873989218328841
Tags
—
Fetched at
Sept. 7, 2026, 9:25 p.m.
Updated at
Sept. 7, 2026, 9:25 p.m.

Description

Every MCP tool call dumps raw data into Claude Code's 200K context window. A Playwright snapshot costs 56 KB, 20 GitHub issues cost 59 KB. After 30 minutes, 40% of your context is gone.I built an MCP server that sits between Claude Code and these outputs. It processes them in sandboxes and only returns summaries. 315 KB becomes 5.4 KB.It supports 10 language runtimes, SQLite FTS5 with BM25 ranking for search, and batch execution. Session time before slowdown goes from ~30 min to ~3 hours.MIT licensed, single command install:/plugin marketplace add mksglu/claude-context-mode/plugin install context-mode@claude-context-modeBenchmarks and source: https://github.com/mksglu/claude-context-modeWould love feedback from anyone hitting context limits in Claude Code.

Enrichment

Theme
utilities for Claude and Claude Code
Vertical
Horizontal
Function
Dev tools
Audience
Developer
AI stance
AI feature
Project type
Hobby / open-source project
Normalized one-liner
compress model context protocol output for claude
Manually corrected
False

Could you build this?

Yes This is a standard MCP proxy middleware that intercepts verbose tool execution payloads, compresses or parses them inside a local container/sandbox, and returns condensed context to the LLM.

Discussion

20 comments analyzed.

Competitors mentioned: Tantivy (BM25 search engine), Context7, Playwright, GitHub MCP servers

Concerns raised: WebFetch hook doesn't trigger reliably (known bug in older versions), Database doesn't persist across sessions, Token counting is estimated (bytes/4), not actual tokenizer, No explicit raw mode toggle for full output, Lossy compression risk—missing information on first query

Feature requests: Add diagram showing data flow and implementation architecture, Persistent cross-session indexing with Tantivy for scalability, Dedicated PII redaction layer (regex-based sensitive data masking), Benchmark performance on smaller models vs. larger context-constrained ones (SWE-bench), Actual token counting instead of byte-based estimation

Competitors

Other products that read as similar to this one — 253 launches clear the similarity bar, closest 8 shown.

Attention rank: #43 of 254 (itself plus its competitors, highest first — normalized so YC and Product Hunt are compared fairly).

Launched 111 days after the earliest competitor.

Other launches for this product

Same idea, different domain

Nobody's really built a dev tools tool for Sales yet.