Skill Claude CodeCodex
Surface and resolve ambiguity before the spec is written: ask targeted questions about underspecified behaviour, edge cases and decisions, and record the answers. Use between proposal and spec whenever the proposal leaves choices open.
Skill Claude CodeCodex
Decide HOW the spec is met: the technical approach, the key architecture decisions and their trade-offs, with Mermaid diagrams where they clarify. Use after the spec is stable, before breaking work into tasks.
Skill Claude CodeCodex
Investigate a feature or bug before committing to an approach: frame the problem, survey the relevant code, and lay out options with trade-offs. Use at the very start of a non-trivial change, when the problem or solution space is still unclear.
Skill Claude CodeCodex
Turn an explored idea into a reviewable proposal: state the why, the scope (in and out), the approach, and the risks. Use once exploration converges and before writing the spec — it is the decision artifact a reviewer signs off.
Skill Claude CodeCodex
Specify WHAT must be true as a delta against the current specs — testable requirements and acceptance scenarios in the OpenSpec delta format. Use after clarify; the spec, not the code, is the source of truth for behaviour. Avoid HOW (that is design).
Skill Claude CodeCodex
Break the spec and design into an ordered, checkable task list — small, verifiable steps that map back to requirements. Use once spec and design are agreed, to drive (and track) implementation.
Skill Claude CodeCodex
Validate the implementation against the spec: walk every requirement and acceptance scenario, run the checks, and report any gaps. Use after apply, before archiving — it is the evidence the change is actually done.
Skill Claude CodeCodex
Create commits per policy (no co-author, with approval, conventional). Trigger: when committing changes.
Skill Claude CodeCodex
Set up VS Code (recommended extensions and profiles) to work well with this workspace. Trigger: when the user asks about extensions, VS Code profiles, or preparing the editor.
Skill Claude CodeCodex
Teaches how to use this AI workspace (SDD, skills, MCP), for people new to AI. Trigger: when the user asks how to use the workspace, how to start, or what SDD is.
Skill Claude CodeCodex
Helps users discover and install agent skills when they ask questions like "how do I do X", "find a skill for X", "is there a skill that can...", or express interest in extending capabilities. This skill should be used when the user is looking for functionality that might exist as an installable skill.
Skill Claude CodeCodex
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
Skill Claude CodeCodex
Create new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy.
Agent Claude Code
Analyze blind comparison results to understand WHY the winner won and generate improvement suggestions.
Agent Claude Code
Compare two outputs WITHOUT knowing which skill produced them.
Agent Claude Code
Evaluate expectations against an execution transcript and outputs.
MCP server Claude CodeCodexCursor
Gives the agent up-to-date documentation and code examples for libraries and frameworks, fetched on demand so answers are not stuck at the model's training cutoff. Runs locally from the @upstash/context7-mcp npm package.
Instructions file CodexOpenCode
Instructions for grojof/ai-workspace-generator, covering ai-workspace-generator — ai agent guide (agents.md), universal conventions (layer 0), encoding & line endings, commits and code style.
Instructions file
Instructions for grojof/ai-workspace-generator, covering ai-workspace-generator — claude code and claude-specific notes.
Skill Claude CodeCodex
Audit the project layout/architecture against the canonical layouts (LAYOUTS + CONV-STRUCT). Trigger: architecture review.
Skill Claude CodeCodex
Audit a REASONS spec/app against a catalogue of security controls (SEC-). Trigger: security review or before approving a spec for deployment.
Skill Claude CodeCodex
Audit the tech stack against the profile's catalogue (TECH-A- / TECH-B-). Trigger: stack review.
Skill Claude CodeCodex
Audit code style/conventions against a catalogue of engineering conventions (CONV-). Trigger: style review.
Skill Claude CodeCodex
Generate application code from a user-reviewed REASONS spec — Profile A (layered FastAPI/SQLAlchemy) or B (ES modules + SRI CDN). Trigger: materialising a signed-off spec as code.