Convert Claude Code or Codex into the most intelligent Deep Research Agent. Collect, search, and synthesize web research into a persistent, searchable wiki that builds on itself. Hosted API + MCP: hyperresearch.ai
Python agentsagentskillscitationsclaude-codecodexcodex-plugincodex-skilldeep-researchdeep-research-agentdeepresearchmcp-serverresearchresearch-agentresearch-tools
prelaunch polish: bigger social wordmark + Python 3.14 disclaimer (#13)
Jordan Gibbs281bd96a88a5
diff --git a/README.md b/README.mdindex 3316296..41e5a2a 100644--- a/README.md+++ b/README.md@@ -28,7 +28,9 @@ pip install hyperresearchhyperresearch install```-Then, in a fresh Claude Code session, type `/hyperresearch <anything>`.+Then, in a fresh Claude Code session, type `/hyperresearch <anything>`.++> **Python 3.11–3.13 only.** 3.14 isn't supported yet — Crawl4AI pins `lxml~=5.3` and that line has no cp314 wheels. Tracking the upstream fix at [unclecode/crawl4ai#1903](https://github.com/unclecode/crawl4ai/issues/1903). Use `pyenv install 3.13`, `uv venv -p 3.13 .venv`, or `py -3.13 -m venv .venv` to create a compatible environment.---diff --git a/assets/_generate_banner.py b/assets/_generate_banner.pyindex a152463..4685119 100644--- a/assets/_generate_banner.py+++ b/assets/_generate_banner.py@@ -31,21 +31,25 @@ src = Image.open(SRC).convert("RGBA")# chunky letterforms instead of blurring them.upscaled = src.resize((src.width * 2, src.height * 2), Image.NEAREST)-# Cap at ~70% of canvas height so the wordmark doesn't crowd the edges-max_h = int(TARGET_H * 0.70)-if upscaled.height > max_h:- ratio = max_h / upscaled.height+# Wordmark should DOMINATE the social card, not float in a sea of dark+# padding. Fill ~92% of the canvas width — leaves a thin margin so the+# letterforms breathe but the content reads clearly when platforms+# crop to tighter aspect ratios than GitHub's native 2:1.+max_w = int(TARGET_W * 0.92)+if upscaled.width != max_w:+ ratio = max_w / upscaled.widthupscaled = upscaled.resize(- (int(upscaled.width * ratio), max_h),+ (max_w, int(upscaled.height * ratio)),Image.NEAREST,)-# Cap at ~85% of canvas width-max_w = int(TARGET_W * 0.85)-if upscaled.width > max_w:- ratio = max_w / upscaled.width+# Sanity cap at 88% of canvas height (mostly defensive — the source+# wordmark's 2.34:1 aspect ratio means width usually constrains first).+max_h = int(TARGET_H * 0.88)+if upscaled.height > max_h:+ ratio = max_h / upscaled.heightupscaled = upscaled.resize(- (max_w, int(upscaled.height * ratio)),+ (int(upscaled.width * ratio), max_h),Image.NEAREST,)diff --git a/assets/banner_social.png b/assets/banner_social.pngindex ec19f66..53ca694 100644Binary files a/assets/banner_social.png and b/assets/banner_social.png differ