jupes/agent-forge-harness / .agents/skills/forgemaster-mini
forgemaster-mini
VerifiedTrimmed Forge pipeline (scope → build → wrap) for small, clear-scope work. Use when the user runs /forgemaster-mini or has a small, well-scoped task.
/forgemaster-mini — Trimmed Forge pipeline for small tasks
Run the low-ceremony Forge path for small, clear-scope work (bugs, chores, config, single-file changes). Same Beads + TDD discipline as the full pipeline, but collapsed into three lightweight steps — scope → build → wrap — with far fewer turns, gates, and artifacts to keep cost down.
Use this directly when you already know the task is small. For anything medium-or-higher, use
/forgemaster (which can also auto-route here).
Usage
Accepts the same inputs as /forgemaster — a JIRA ticket, a Beads issue id, or a
free-text description.
/forgemaster-mini <task description> # free text
/forgemaster-mini <BEADS-ID> # e.g. agent-forge-harness-f25 → run mini against that issue
/forgemaster-mini <JIRA-KEY> # e.g. PROJ-1234 → mirror into Beads, then run mini
/forgemaster-mini <task> --smith <name> # also record the run, as built by a configured smith
What to do
Follow .claude/workflows/forge-mini.md in full. In short:
- Preflight —
bd ready >/dev/null; if it errors,bd dolt startand retry. Beads is required. - Scope — read only the touched files; ask at most one decision question (lead with the
plain-language why, then options); state the approach inline. Resolve the input to one Beads
task:
bd showan existing id, mirror a JIRA key into Beads (--external-ref jira-<KEY>, pasting the ticket if no JIRA integration is configured), or create a new task for free text — then--claimit (priority via.claude/skills/beads-priority-assignment/SKILL.md). Confirm the approach once. - Build — TDD via
.claude/skills/tdd/SKILL.md(vertical slices) for code; one runnable demo;worklog:comment; commit with tests. - Wrap — quality gates, close the task with evidence, push (+ PR if warranted; the tracker is local-only, so no
bd dolt push), and report inline: what changed, how to verify, Beads id + PR link.
No plans/ or reports/ doc files and no forge:phase-gate / run state file, unless it was given
--smith — tracking lives in Beads. Escalate to the full pipeline (/forge-research <slug>) if
the work outgrows "mini".
--smith <name>
The mini path keeps no run state, so a --smith is recorded by one write at the very end. Pick a
kebab-case <slug> for the run, and after the wrap step's push:
- Save the handoff you report as
reports/<slug>-ship.md. bun run forge:phase-gate ship --slug <slug> --write --bead <task-id> --smith <name>
That call checks the name (a smith that is not configured is refused, with the list of the ones that
are: correct it and run the call again; nothing is lost, because this session builds either way),
stores the smith on the run as its executor, and leaves a run that is already shipped, so it never
shows as in flight. --smith does not change the model of the session reading this command. A mini
run has no linked gate entry, with or without --smith: its checks are the bare
bun run typecheck && bun run lint && bun test of the wrap step.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deployment | Stars: 318k | Updated | Claude CodeCodex | Verified | Install |
| xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved. | Stars: 196k | Updated | Claude CodeCodex | Verified | Install |
| shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy. | Stars: 118k | Updated | Claude CodeCodex | Verified | Install |
| childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release. | Stars: 117k | Updated | Claude CodeCodex | Verified | Install |