agent-spec-creation
Agent Spec Creation Workflow
Purpose
Create implementation-ready agent integration specifications without changing production code.
When to Use
Use this skill when the task is to define or revise agent behavior, inputs, outputs, error handling, or acceptance criteria in specs.
Required Workflow
- Read intent and existing spec context first:
specs/README.mdspecs/agents.md- relevant files in
specs/agents/*.md
- Rely on the
spec-creatorskill for document structure and quality requirements. - Study the target CLI behavior before writing spec details:
<agent-cli> --help- relevant subcommand help (for example
<agent-cli> run --help)
- Define explicit, testable requirements:
- command shape and required flags
- environment variables
- output parsing/streaming expectations
- non-interactive execution behavior
- failure modes and expected surfaced errors
- Update spec files only.
Constraints
- Do not implement code in
internal/agent/*while using this skill. - Do not run iterative Ralph loop automation.
- Keep the workflow linear: read, specify, and stop after spec updates.
Completion Criteria
- Spec changes are clear, deterministic, and testable.
- Scope and non-goals are explicit.
- Related index files (for example
specs/README.md) are updated when new spec files are added.
More from iyaki/ralph
spec-creator
Generate comprehensive technical specifications for usage with Ralph.
10code-search
Use when you need to locate definitions of classes, functions, constants, types, or symbols in the application code. Before searching individual files, run Repomix with stdout output to capture a snapshot and base the search on that content.
1shell-command
Execute shell commands with strict adherence to non-interactive flags and environment variables to prevent hanging in OpenCode's headless environment.
1skill-creator
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.
1test-driven-development
Use when implementing any feature or bugfix, before writing implementation code
1agent-implementation
Implement or update support for an underlying agent CLI in Ralph after specs are defined, including factory wiring and tests.
1