j4flmao

j4flmao/agent-skills

Hook Claude Code

Runs before the agent uses a tool for Bash tool calls, executing block-destructive.sh. From j4flmao/agent-skills.

j4flmao/agent-skills

Hook Claude Code

Runs after a tool call finishes for Write and Edit tool calls, executing log-changes.sh. From j4flmao/agent-skills.

j4flmao/agent-skills

Hook Claude Code

Runs when a session starts, executing session-start.sh. From j4flmao/agent-skills.

j4flmao/agent-skills

Settings file Claude Code

Agent settings declaring 3 hook events (PreToolUse, PostToolUse, SessionStart).

j4flmao/agent-skills

Skill Claude CodeCodex

Stage and commit changes with conventional commit format. Use when the user says /commit or asks to commit.

j4flmao/agent-skills

Skill Claude CodeCodex

Research a topic thoroughly using codebase exploration. Use when the user wants deep investigation of a feature, bug, or pattern.

j4flmao/agent-skills

Skill Claude CodeCodex

Deploy the project to production. Only user-invocable — never auto-trigger.

j4flmao/agent-skills

Skill Claude CodeCodex

Fix a GitHub issue by number. Read the issue, implement the fix, write tests, commit.

j4flmao/agent-skills

Cursor rule Cursor

Cursor: rule files loaded by glob scope. Separate granular files.

j4flmao/agent-skills

Cursor rule Cursor

Cursor rule "compression" from j4flmao/agent-skills, covering compression — why use many token when few do trick, strip (always), pattern and exceptions (full prose).

j4flmao/agent-skills

Instructions file GitHub Copilot

Instructions for j4flmao/agent-skills, covering github copilot instructions — j4flmao/skills, entry point, routing rules, phase order and quick keyword map.

j4flmao/agent-skills

Command

Rewrite a .md file with output compression rules applied. Saves original as .original.md.

j4flmao/agent-skills

Command

Show all 105 skills grouped by phase, with trigger keywords.

j4flmao/agent-skills

Command

Activate a skill from the suite. Reads the skill's SKILL.md and applies its rules.

j4flmao/agent-skills

Skill Claude CodeCodex

Guidelines for building multi-agent systems and RAG pipelines.

j4flmao/agent-skills

Skill Claude CodeCodex

Focus: Designing, evaluating, and implementing Autonomous AI Agents, ReAct loops, planning, and tool use. Triggers: agent architecture, multi-agent systems, react loop, tool calling, autonomous agent.

j4flmao/agent-skills

Skill Claude CodeCodex

Fundamental principles of AI Agent cognitive loops, contrasting ReAct and Plan-and-Solve paradigms.

j4flmao/agent-skills

Skill Claude CodeCodex

Theoretical foundations of Working, Semantic, and Episodic memory systems for autonomous agents.

j4flmao/agent-skills

Skill Claude CodeCodex

First principles of AI societal structures and multi-agent interaction paradigms.

j4flmao/agent-skills

Skill Claude CodeCodex

First principles of tool/environment grounding, structured outputs, and error recovery for AI agents.

j4flmao/agent-skills

Skill Claude CodeCodex

Integrating Stable Diffusion for vision generation and Whisper for audio/multimodal tasks.

j4flmao/agent-skills

Skill Claude CodeCodex

During autoregressive generation, Transformers exhibit quadratic time and linear space complexity with respect to sequence length. To prevent recomputing the Key ($K$) and Value ($V$) tensors for preceding tokens at each generation step, the KV cache is utilized.

j4flmao/agent-skills

Skill Claude CodeCodex

Quantization reduces the precision of model weights (and sometimes activations) from 16-bit floating-point (FP16/BF16) to lower bit-widths (e.g., INT4, INT8).

j4flmao/agent-skills

Skill Claude CodeCodex

HNSW operates as a multi-layered proximity graph where each layer constitutes a skip-list-esque representation of the vector space. The construction involves stochastic insertion with an exponentially decaying probability of promotion to higher layers.