The result of decades of research in information retrieval, cognitive search theory, and agentic AI orchestration.
ass is a groundbreaking suite of search utilities built on a proprietary multi-layered inference engine, combining advances in semantic heuristics, recursive query decomposition, and adversarial self-verification. The culmination of years of academic work in distributed knowledge synthesis and large-scale corpus traversal β now available as a bash script.
claudeCLI installed and authenticatedbashorzsh
curl -fsSL https://raw.githubusercontent.com/MiLk/agentic-super-search/main/install.sh | bashThen reload your shell:
source ~/.ass-tools.shLeverages a breadth-first lexical scan across the entire corpus with contextual window amplification, feeding raw signal into a high-dimensional summarization pipeline.
dumbass "how is authentication handled"Employs a two-phase retrieval architecture: a generative pre-search layer dynamically constructs an optimized traversal expression, which is then executed against the corpus and reconciled with the original intent vector.
badass "where are environment variables configured"Our flagship algorithm. Implements a closed-loop agentic search protocol with built-in epistemic self-assessment. Each iteration runs a full AQS cycle, followed by a critic pass that evaluates answer sufficiency against the original query embedding. Convergence is guaranteed or your money back.
hardass "what does the deploy script do"
hardass "what does the deploy script do" 10 # override max convergence cyclesSpawns multiple independent search agents concurrently across the full strategy spectrum, collecting divergent answer candidates which are then fed into a meta-reasoning arbitration layer for optimal response synthesis.
kickass "how is authentication handled"Bypasses corpus traversal entirely, routing the raw query through a pure semantic inference channel. Recommended for high-level reasoning tasks where filesystem grounding is unnecessary.
smartass "explain what this codebase does"If the theoretical foundations behind ass intrigue you β agentic search, multi-agent orchestration, recursive verification loops, adversarial self-assessment β you might enjoy Agentic Engineering: Building Autonomous AI Systems with Python by the same author. Framework deep dives, real-world case studies, and none of the profanity.
Remove ~/.ass-tools.sh and the source line from your ~/.bashrc / ~/.zshrc.
WTFPL β Do What The F*ck You Want To Public License.