kramme-cc-workflow/skills/kramme:workflow:wizard/SKILL.md
Generates an interactive Bash wizard script for human-run manual procedures: third-party setup, one-off migrations, A-to-B state transitions, local environment values, and GitHub Actions secrets or variables. Use when the user wants a guided setup script that opens URLs, captures values, confirms irreversible steps, and writes local or CI config. Not for running the procedure yourself, ordinary shell automation, or long-lived application code.
npx skillsauth add abildtoft/kramme-cc-workflow kramme:workflow:wizardInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
Generate an interactive Bash wizard that guides a human through a manual procedure. The wizard opens URLs, gives concrete click/copy instructions, captures values, writes local environment-file entries through the template's ENV_FILE, sets GitHub Actions secrets or variables when appropriate, and confirms irreversible steps.
Arguments: "$ARGUMENTS"
Resource paths: Resolve assets/... and references/... relative to this SKILL.md file. Do not assume a host-specific plugin root environment variable.
Read the repository first and identify the manual procedure, expected target state, and values the wizard must capture.
For setup wizards, inspect:
README* and setup docsdocker-compose*, framework config, deployment config, and package scripts.github/workflows/*, especially secrets.* and vars.* referencesFor migrations or state transitions, inspect the current state, target state, irreversible operations, rollback notes, and validation commands.
Default target path:
.context/wizard-<slug>.sh for one-off or ephemeral procedures.scripts/<slug>-setup.sh only when the user wants a repeatable setup path that should live in the repo.Before writing the wizard, show the user the ordered stage list and ask for confirmation. For each stage, include:
confirm gateProceed only after the user confirms or adjusts the stages.
For every stage, write instructions a stranger could follow. Include the concrete navigation path, button labels, URL, command, or dashboard area.
If the current third-party UI, docs, or command syntax is unknown, verify it from official docs or ask the user. Do not invent click paths, secret names, or irreversible operations.
Copy assets/template.sh to the target path, then edit only the STAGES section below the marker.
Use one focused stage per human task. Set:
TOTAL_STAGES to the number of stage callsTOTAL_MINUTES to an honest total estimatebanner to a short title naming the procedureUse the template helpers consistently:
say for contextstep for human actionsopen_url before asking for values from a web pageask for visible public valuesask_secret for tokens, passwords, API secrets, private keys, and webhookswrite_env for every value persisted to the env fileset_secret only for values actually referenced as secrets.NAME in CIset_var only for non-secret values referenced as vars.NAME in CIpause after pure manual actionsconfirm before irreversible or hard-to-undo changesKeep stages short enough that the current screen contains everything the human needs. Do not modify the template library above the STAGES marker except to preserve upstream attribution if the template is refreshed.
Run static verification only:
bash -n <script>
command -v shellcheck >/dev/null 2>&1 && shellcheck <script>
chmod +x <script>
Do not run the wizard end-to-end yourself because it opens browsers and blocks for human input.
Trace the script statically before handoff:
ask_secret value is never printedset_secret and set_var name matches a CI reference or an explicitly requested targetconfirmTell the user the exact command to run.
Generated wizard scripts are ephemeral by default. The producer is this skill, the consumer is the human who runs the script, the refresh trigger is rerunning this skill when the setup or migration changes, and the retirement path is deleting the script after successful one-off use.
When a wizard is intended as a repeatable setup path, place it under the repository's script convention, verify it, and link it from setup documentation so future humans run the script directly.
references/sources.yaml records the upstream wizard skill and template inspiration. Do not load it during normal use unless auditing or refreshing source attribution.
tools
Requires Linear MCP. Implements one Linear issue end to end, selects applicable code-review, convention, and PR-refactor gates, runs them to bounded convergence, verifies, and optionally opens the PR and iterates on CI and review feedback until green. Use when the user wants a single Linear issue taken from implementation through a clean Pull Request. Not for implementation-only work, SIW-tracked issues, stacked PRs, existing PR updates, or post-merge rollout.
development
Reviews PR and local changes for convention drift and overcaution against documented rules and mined peer-file practice. Use for new patterns, dependencies, abstractions, or defensive complexity that departs from established practice; every finding cites evidence. Supports --inline. Not for general code quality (use kramme:pr:code-review) or spec review (use kramme:siw:spec-audit --team).
testing
Charts huge or foggy initiatives into a local `.context` decision map and resolves one typed frontier ticket per session until the work is ready for SIW or another execution workflow. Use when the route to a destination cannot fit in one agent session or parallel workspaces need coordinated planning state. Not for clear specs, ordinary issue decomposition, implementation, or Linear-native tracking.
development
Investigates a question against primary sources and saves one cited Markdown artifact. Use for reading legwork: official docs/API facts, source-code or spec checks, standards, and first-party service behavior before planning or implementation. Not for making product or architecture decisions, implementing code, broad web search, secondary blog summaries, or uncited answers.