From rf-agentskills
Generates Robot Framework user keywords from JSON input with arguments, documentation, tags, setup/teardown, steps, and returns. Detects embedded-argument style from projects.
How this skill is triggered — by the user, by Claude, or both
Slash command
/rf-agentskills:keyword-builderThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Create user keywords in Robot Framework syntax from structured input. Output JSON only.
Create user keywords in Robot Framework syntax from structured input. Output JSON only.
Provide input via --input or stdin. Example:
{
"keyword_name": "Create User",
"description": "Creates a new user via the UI.",
"arguments": [
{"name": "username", "type": "str"},
{"name": "role", "default": "viewer"}
],
"tags": ["ui", "smoke"],
"setup": {"keyword": "Open Browser", "args": ["${URL}", "chromium"]},
"teardown": {"keyword": "Close Browser"},
"style": "simple",
"steps": [
{"keyword": "Click", "args": ["Add User"]},
{"keyword": "Input Text", "args": ["Username", "${username}"]},
{"keyword": "Click", "args": ["Save"]}
]
}
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/keyword_builder.py" --input keyword.json
Detect embedded-argument style from an existing project:
python3 "${CLAUDE_PLUGIN_ROOT}/scripts/keyword_builder.py" --input keyword.json --project-root . --detect-embedded
Use "return_value" to add a RETURN statement at the end of the keyword:
{
"keyword_name": "Get Full Name",
"arguments": [{"name": "first"}, {"name": "last"}],
"steps": [{"assign": ["${result}"], "keyword": "Catenate", "args": ["${first}", "${last}"]}],
"return_value": "${result}"
}
Multiple return values:
"return_value": ["${var1}", "${var2}"]
[Arguments] using ${name}: type syntax (e.g. ${count}: int). This builder
does not yet generate that syntax; type information is recorded in [Documentation]
only.artifact: keyword blockwarnings and suggestionsmeta: any detected project conventionsnpx claudepluginhub manykarim/robotframework-agentskills --plugin rf-agentskillsGenerates Robot Framework test cases from structured JSON inputs for requirements or scenarios. Supports keyword-driven/template styles, tags, setup/teardown, timeouts, and control structures.
Expert approach to keyword-driven-testing in test automation. Use when working with .
Generates structured PROMPT.md files for autonomous AI workflows (Ralph Loops) from rough descriptions or PRDs, with phases, validation criteria, completion promises, and blind review.