🎬 DEMO SHOWCASE — GSOC-CONTRIB v0.5.3 | ZERO COLD-START WORKSPACE ENGINE
HAILUO_AI_BOOTH_PROMO_4K.MP4
BOOTH DISPLAY LOOP
$ contrib start
Avg Setup Speed3.4s⚡
Bandwidth Saved95.4%
JustVugg/colibri50% .git saved
AI Discovery Latency80% Skipped
WE ARE LIVE ON NPM & PYPI|gsoc-contrib v0.5.3

SPARSE. BLOBLESS.

CONTRIBUTE.

contrib-cli is the zero-clone workspace engine between GitHub issues and your IDE. Sub-3.4s workspace creation. 95% bandwidth savings. Instant developer onboarding.

// SECTION: AI_AGENT_CASE_STUDY
ANTIGRAVITY IDE VERIFIED007
ZERO COLD-START LATENCY FOR AI AGENTS

STRUCTURED AI BRIEFING ENGINE

When you launch a workspace with $ contrib start <url> --agy, contrib-cli deterministically generates a 3-file briefing package inside .contrib/. This eliminates 80% of codebase discovery friction for AI Pair Programmers.

GENERATED BRIEFING PACKAGE (.contrib/)
context.json
METADATA

Machine-readable metadata — repo, issue #, branch, stack, test/build commands, quality tools

ISSUE.md
CONTEXT

Raw issue content, reproduction steps, linked discussions, and target labels

AI_PROMPT.md
SYNTHESIS

Synthesized AI prompt — combines issue body, focus area code paths, CONTRIBUTING.md, PR template, & rules

Skips 80% of AI codebase discovery time
Pre-creates feature branch (contrib/issue-934)
AI AGENT ANALYSIS
// ANTIGRAVITY IDE OBSERVED:

“That's a very well-designed tool. It takes a GitHub issue and produces a structured AI briefing package in .contrib/.”

“The focus_areas extraction pulls specific code references out of the issue body and surfaces them as entry points. That's what let me skip 80% of discovery.”

“Contributing to open source has enormous cold-start cost for AI agents. .contrib eliminates the cold start by doing that translation once per issue, offline, deterministically.”

// LIVE BENCHMARK: JustVugg/colibri50% .GIT SAVINGS
Full Clone .git:17.67 MB
contrib Workspace .git:8.86 MB (50% smaller⚡)
Verified in Antigravity IDEIssue #934 Benchmark
// SECTION: WHAT_WE_ARE_BUILDING
MISSION ARCHITECTURE003

WHAT WE ARE BUILDING

contrib-cli is engineered to bridge open-source issue discovery, instant zero-clone workspaces, and modern IDE workflows into a single seamless CLI.

CORE ENGINE
[01]

ZERO-CLONE WORKSPACE ENGINE

Sub-3.4s workspace initialization leveraging Git blobless clones (--filter=blob:none) and sparse checkouts. Stop downloading multi-gigabyte repo histories to fix a single line of code.

95.4% average bandwidth reduction
91.2% local disk space savings
Automatic branch isolation per issue/PR
Instant workspace status tracking & cleanup
$ contrib start https://github.com/org/repo/issues/101
SMART MATCHING
[02]

AI GSOC & ISSUE RECOMMENDER

AI-powered recommendation engine matching contributor skill sets (languages, domains, experience level) with official GSoC organizations and open contribution opportunities.

Interactive quiz mode (contrib recommend)
Skill filtering (--lang python,js --domain ai)
Good-first-issue discoverability across 100+ orgs
Historical GSoC repo indexing & classification
$ contrib recommend --lang python,javascript --level beginner
NATIVE INTEGRATION
[03]

ANTIGRAVITY IDE & EDITOR BRIDGE

First-class editor launch flags. Open your newly created lightweight workspace instantly in Antigravity IDE (--agy), Neovim, Vim, Helix, Zed, VS Code, or Cursor.

Native --agy / --ide / --antigravity flags out of the box
Terminal editor auto-detection (Neovim, Vim, Helix)
One-command open for active workspaces (contrib open --agy)
Cross-platform support (Windows, macOS, Linux)
$ contrib start https://github.com/org/repo/issues/101 --agy
METADATA ENGINE
[04]

PRE-CLONE ISSUE ANALYZER

Inspect GitHub issue context, linked pull requests, labels, and relevant target repository file paths directly from your terminal before cloning a single file.

GitHub REST & GraphQL API metadata resolution
Target file impact estimation
Linked PR status and discussion inspection
Offline workspace state caching
$ contrib analyze https://github.com/org/repo/issues/101