HARK
Execution memory for Agent Skills: scoped recall, run evidence, recovery paths, and invalidation for a query-regression workflow.
Public work / GitHub
Built to explore.
Open to inspect.
A collection of public projects and experiments across AI, software, and product engineering. The source makes the decisions—and the scope—visible.
Featured work
View my GitHubLatch
A dataset release-review workflow combining source context, measured checks, human approval, and hash-verifiable release evidence.
GridForge
Off-grid solar design that normalizes web component inventory and checks panel, inverter, and battery combinations against electrical constraints.
DO-SIFT
Digital-forensics workflow orchestration with approved adapters, execution guardrails, an evidence ledger, and claim validation.
fenoa
A creative application for fictional worlds and remix branches, with versioned content and shared domain operations for creators and agents.
ShipSafe
GitLab review orchestration connecting trusted verdict notes to merge checks, issues, and generated fix merge requests.
All public repositories
The full collection, from focused systems to early experiments.
90 repositories · Original projects lead; forks are explicitly attributed.
HARK
Execution memory for Agent Skills: scoped recall, run evidence, recovery paths, and invalidation for a query-regression workflow.
Latch
A dataset release-review workflow combining source context, measured checks, human approval, and hash-verifiable release evidence.
GridForge
Off-grid solar design that normalizes web component inventory and checks panel, inverter, and battery combinations against electrical constraints.
DO-SIFT
Digital-forensics workflow orchestration with approved adapters, execution guardrails, an evidence ledger, and claim validation.
fenoa
A creative application for fictional worlds and remix branches, with versioned content and shared domain operations for creators and agents.
ShipSafe
GitLab review orchestration connecting trusted verdict notes to merge checks, issues, and generated fix merge requests.
MoveScore
A choreography-to-music application combining video analysis, musical direction, and generated audio through a Python backend.
airstrong
An airline-disruption recovery prototype with candidate planning, an operational twin, and approval before operational changes.
PrathmeshAdsod
The public GitHub profile README introducing interests and selected projects.
Mirra
A fashion-campaign virtual-mirror implementation with brand review, published manifests, and durable try-on provider tasks.
SKILL.md-Collections
A collection of agent skills and helper scripts with library validation and fixture-based checks.
RedDefence
Reddit moderation tooling with classification, duplicate detection, policy modes, and audit-oriented moderation workflows.
STILL-Scriptures
A video-reflection application with timestamp-aware analysis, viewing progress, and source-backed Scripture references.
PAINT-IT
A Reddit memory game with timed doodle recall, creator-authored challenges, and aggregated community reconstructions.
Red-Eagle
A supply-chain policy prototype separating model-assisted policy interpretation from governed inventory and transfer decisions.
WhyRight
A classroom-reasoning application for capturing explanations, exploring misconceptions, and comparing revised answers.
Slacktory
A Slack workflow-discovery prototype with blueprint generation, guided approvals, and configurable ticket actions.
Supplychain-Long-Term-Memory
A local supply-chain decision prototype with a traceable LangGraph workflow and approval steps.
Public visibility verified 8 October 2026. Descriptions reflect inspected repository scope; a source listing does not imply a launched product. Sort dates represent repository pushes. Explore on GitHub ↗