Skill v1.0.0
currentLLM-judged scan90/100version: "1.0.0" name: map description: Map an existing codebase (stack and commands, structure, conventions, tests, risks) so VBW plans fit the code; saves .vbw/map.md. allowed-tools: Bash("${CLAUDE_PLUGIN_ROOT}/bin/vbw" ) Bash(vbw ) Workflow(vbw:mapping)
"${CLAUDE_PLUGIN_ROOT}/bin/vbw" config models 2>&1 || true"${CLAUDE_PLUGIN_ROOT}/bin/vbw" config effort 2>&1 || true"${CLAUDE_PLUGIN_ROOT}/bin/vbw" workflows on 2>&1 || true"${CLAUDE_PLUGIN_ROOT}/bin/vbw" interview 2>&1 || true
Write to the user's level, explanation depth and involvement (above; "-": plain words), and pass them to the workflow as "profile": {"level", "depth", "involvement"} (none for "-").
(Outside a VBW project the effort line shows an error: omit effort.)
If this is not a VBW project yet ("not a VBW project" above), set it up first: vbw init and vbw statusline on. VBW needs the Workflow tool: the workflows line above turns it on (say so in one line when it did); without it, say why and stop.
Run VBW_SESSION_ID=${CLAUDE_SESSION_ID} vbw run start map, then start the Workflow vbw:mapping with args {"models": <the JSON above>, "effort": <the effort JSON above>, "profile": ...}. When it returns, VBW_SESSION_ID=${CLAUDE_SESSION_ID} vbw run end, then write its map to .vbw/map.md (replacing any older map), then tell the user in a few lines what the codebase is and how to build and test it, and name any angle listed in missing. The Architect and Lead read this map when they plan.