Scaffolds a complete, runnable Spring Boot application on top of the flow-driven-domain (FDD) library (io.github.progmodek:flow). Turns a natural-language process description into states/actions/transitions, a workflow JSON, and the full project.
Architecture-agnostic interpretability with nnterp on nnsight 0.8 - one module vocabulary and standard values (layeroutput, attentionoutput, attentionprobabilities, ...) across 98 transformer families, the vision side of vision-language wrappers (model.vision, the tower blocks, the projector, imagetokenmask, image.
★not rated 15▲
+1 3d agoA
tokens not measured
originalMIT
Build and run internal AI agents with Hubzoid from Claude Code. Turn your task, material and rules into a working Hub, test it, save a correction and share it with your team.
A collection of diagnosis-first tutoring skills for explanations, guided problem solving, mistake review, vocabulary practice, recall testing, and study planning.
★not rated 15 11d agoA
tokens not measured
originalMIT
(Alpha) Persistent memory, architectural decisions, and safety guardrails for Claude Code. Your agent starts every session with full project context — stack, decisions, patterns, safety rules, and a handoff from the previous session.
★not rated 14 7d agoA
tokens not measured
originalMIT
Answer at the decision level - plain language you can judge and decide on, with code, paths and line numbers kept below a marked line. Armed per session with /decision-layer, off by default.
★not rated 14
changed 2mo agoA
tokens not measured
originalMIT
CodeEraser passive guard + erase skill: a winnowing clone index behind a per-project daemon intercepts T1/T2 duplicate writes at PreToolUse (with the tombstone and flow legs judged in the core), records at PostToolUse whether an asked write ran (settle), audits net LOC + touched duplicates at Stop, and reports health…
Force a human review gate after every Claude turn that edits files. Summarizes what changed and why, drops one-line doc comments on touched classes and functions, then waits for the developer to Approve or Undo.
Bootstrap or adopt a project with a lean AI-Human collaboration framework: a concise CLAUDE.md, architecture and plan docs, enforcement hooks, and tag-based versioning.
★not rated 14
changed 10d agoA
tokens not measured
originalMIT
A plugin containing Chinese-language skills for saving information to a wiki, producing AI news reports, analysing observations, creating content and courses, reviewing growth and paid content, formatting Markdown, and handling Git commits and pushes.
★not rated 14 20d agoA
tokens not measured
originalMIT
A senior product-marketing and growth teammate with compact capabilities for research, positioning, campaigns, creative, copy, social, search, conversion, measurement, and Vietnamese.
★not rated 14 8d agoA
tokens not measured
originalMIT
Fully autonomous development assistant that handles all technical tasks including git operations without asking for permission. Use for feature implementation, bug fixes, refactoring, and code changes that need commits.
★not rated 14 7mo agoA
tokens not measured
originalMIT
MCP action building blocks - Replay tool sequences to save tokens. Use for: (1) Full automation when all steps are deterministic, (2) Micro patterns like click-and-wait or scroll-and-wait where Sandy executes actions, then LLM calls screenshot for visual analysis.
★not rated 14 9mo agoA
tokens not measured
originalApache-2.0
Inventory the AI-agent skills and instruction files installed on this machine. Scans where Claude Code, Codex, Cursor, Copilot, Windsurf, and other agents store skills and standing instructions, then reports what is actually there.
★not rated 14 2mo agoA
tokens not measured
originalApache-2.0
Operate renef (renef.io) — the Android ARM64 dynamic-instrumentation engine. Bundles the 'renef' Agent Skill: an operator's manual (syntax contract, self-lint, task playbook) for hooking native & Java functions, scanning/patching memory, tracing syscalls, bypassing SSL pinning / root detection, and porting Frida /…
★not rated 14 4mo agoA
tokens not measured
originalApache-2.0
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: