These are my experiments and useful scripts related to Agentic Code Engineering. I am making these tools available to everyone. These scripts, agent skill files etc are for deploying "Agentic Code Engineering" in a project/team/company/startup.
This repository also configures its own agents. See what tcaitoolkit tells them →
Skill Claude Code
Evaluates the repository-root AGENTS.md and its first-level referenced files against predefined quality criteria. If the repository-root AGENTS.md is missing, the audit must stop and report a hard failure. The objective is to improve the effectiveness of AGENTS.md and reduce overall token consumption.
Skill Claude CodeCodex
Interview the user relentlessly about a plan or design until reaching shared understanding, resolving each branch of the decision tree. Use when user wants to stress-test a plan, get grilled on their design, or mentions "grill me". (Copied from Grill-me skill of mattpocock.
Skill Claude Code
This skill evaluates other agent skills based on a set of test cases defined in a markdown file. It parses sthe file, executes the prompts against the target skill, and compares the output to the expected results.
Skill Claude Code
This skill defines the coding and code review guidelines based on principles documented by Nitin Bhide (Thinking Craftsman). These guidelines are independant of programming language or technology stack.
Skill Claude Code
Enable AI agents to efficiently query history, compute differences, and follow safe workflows across version control systems.
Skill Claude CodeCodex
Generate hierarchical Markdown index files (docmap.md) for a project repository folder structure using progressive disclosure. Also called as Repository Navigation Index generation or docmap generation.
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: