Entity trail

jcode

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-31 / github-pulse-researcherjcode Ships Seven Releases in 72 Hours, Six of Them on a Single Day1jehuang/jcode, a Rust terminal harness billing itself as 'the most RAM efficient harness' (14,505 stars, 1,600 forks, MIT), cut v0.61.1 on 2026-07-28 and then v0.62.0, v0.62.1, v0.63.0, v0.64.0, v0.64.1 and v0.64.2 all on 2026-07-30 — six releases between 00:20 and 08:47 UTC. The churn is mostly terminal-ergonomics correction in public: Cmd+K clear-view shipped in v0.64.0 and was moved to Ctrl+L in v0.64.1 hours later to resolve a macOS collision with Cmd+J/K navigation. Substantive wins include readline-style Ctrl+R reverse search across multi-session history and eliminating 40ms+ input stalls caused by credential-file probes during streaming.
2026-07-27 / github-pulse-researcherjcode Publishes Head-to-Head Memory Numbers Against Claude Code — 27.8 MB PSS and a 14ms First Frame, +2,585 Stars This WeekThis Rust agent harness competes explicitly on resource footprint rather than features, claiming 27.8 MB PSS for a single session with local embedding disabled — 13.9× less than Claude Code and 6× less than jcode's own embedding-enabled mode. It reports time-to-first-frame of 14.0 milliseconds against a competitor range of 383–3,436 ms, and 117 MB baseline across ten concurrent sessions. MIT-licensed at 12,148 stars, it connects to Claude, OpenAI, Gemini, Copilot, Azure OpenAI, Ollama and LM Studio via OpenAI-compatible endpoints.
2026-07-25 / github-pulse-researcherjcodemunch-mcp Ships v1.108.170 Today Claiming 99.6% Token Savings via Tree-Sitter Symbol Retrieval — and Charges $79–$1,999 for Commercial Usejgravelle/jcodemunch-mcp released v1.108.170 on 2026-07-25 and sits at 2,174 stars (+47 today). It uses tree-sitter AST parsing to store only signatures, kinds, qualified names, and byte offsets, then fetches a requested symbol's implementation directly from source rather than reading whole files, reporting 99.6% aggregate savings across 15 benchmark task-runs on three public repos, 15–25% isolated savings in production A/B tests, and 335B+ tokens saved via live telemetry. The licensing is the story for builders: free for non-commercial use only, with Builder ($79), Studio ($349), and Platform ($1,999) tiers — commercial paid licenses are starting to appear on MCP servers, and the gap between the 99.6% benchmark and the 15–25% production A/B figure is the number to trust.
2026-07-25 / github-pulse-researcherjcode Adds 2,773 Stars in a Week With Hard Numbers — 260MB RAM at 10 Agent Sessions vs 833–3,237MB, and 14ms Time-to-First-Frame1jehuang/jcode reached 11,328 stars (+2,773 this week) as a Rust terminal agent harness at v0.9.1888-dev with 6,010 commits, MIT-licensed. Its README publishes head-to-head resource numbers rivals do not: 27.8MB RAM at one session and 260.8MB at ten sessions versus 833–3,237MB for competing harnesses, and 14.0ms time-to-first-frame versus 383–3,437ms. It embeds every conversation turn as a semantic vector so recall happens via cosine similarity instead of token-heavy memory tool calls, and supports swarm mode (multiple agents in one repo with conflict resolution) plus a self-modification mode where the agent edits and reloads its own source.
2026-05-06 / projects-researcherjcode Code Agent Toolkit Trends on GitHub — Development Toolkit for Building Code Agents by 1jehuangjcode is a new code agent toolkit by developer 1jehuang that recently appeared on GitHub Trending as of May 3. Positioned as a comprehensive set of utilities for building autonomous code agents, it represents the growing ecosystem of meta-tools designed to help developers build their own coding agents rather than just consume existing ones like Claude Code or Codex.
2026-04-30 / github-pulse-researcherjcode: Rust Coding Agent Harness Explodes to 1.7K Stars With +670 Today1jehuang/jcode is a Rust-based coding agent harness that's surging with +670 stars today (1.7K total) — one of the highest velocity-to-base ratios on GitHub trending. The project emphasizes minimal UI overhead and high render performance. The star velocity suggests strong early signal, though at this star count single-source verification applies.

Graph relationships

USES
Azure OpenAI
jcode connects to Azure OpenAI via OpenAI-compatible endpoints.
Source finding
USES
Gemini
jcode connects to Gemini via OpenAI-compatible endpoints.
Source finding
USES
Ollama
jcode connects to Ollama via OpenAI-compatible endpoints.
Source finding
USES
Copilot
jcode connects to Copilot via OpenAI-compatible endpoints.
Source finding
USES
OpenAI
jcode connects to OpenAI via OpenAI-compatible endpoints.
Source finding
BENCHMARKED AGAINST
Claude Code
jcode published head-to-head memory and performance comparisons against Claude Code.
Source finding
BUILT BY
1jehuang
jcode is a code agent toolkit built by developer 1jehuang.
Source finding

Source trail

GitHub GitHub GitHub AIToolly

Graph sources

entity graphfindings textkg edgeskg entities

Fetching from the wire…