tokenbender/agent-guides
one page guides that i let my subscribed/customised agents consume to perform actions
What it solves
This project provides a comprehensive set of guides and a production-ready collection of skills for Claude Code, helping users move from basic usage to advanced agentic workflows. It addresses the gap between raw model capabilities and the same-day production deployment of complex tasks like adversarial cross-pollination, SFT data auditing, and browser automation.
How it works
The repository is structured as a set of documentation guides and a library of reusable "skills" (using the SKILL.md standard). These skills are essentially modular instructions and scripts that can be installed into the .claude/skills/ directory to extend the agent's capabilities. The project also includes scripts for exporting session transcripts for analysis.
Who it’s for
Developers and power users of Claude Code who want to implement advanced orchestration patterns, custom commands, and deterministic guardrails (hooks) to automate complex technical workflows.
Highlights
- Production Skill Collection: Includes 13 specialized skills, such as
multi-mindfor adversarial judgment andcapability-horizon-estimatorfor predicting model success rates. - Advanced Orchestration: Guides on using subagents with worktree isolation, lifecycle hooks for deterministic guardrails, and path-scoped memory.
- Tooling: A Python script to extract and clean Claude session transcripts from JSONL format to Markdown.
- Standardized Layout: Adheres to the Agent Skills open standard for dynamic context injection and distribution.
Related
- Project
- Project
- Project
- Project
- Dispatch