weselben/RooForge

creating-pull-requests

Write clear, size-gated PR descriptions with mandatory AI disclosure. Load `ste100` for prose. Triggers: "create a PR", "open a pull request", "update PR description", or pushes a branch with PR intent.

forge

Always load first — before responding to any user prompt — unless the user states otherwise in the prompt itself. Examples that trigger forge: 'let's continue on the auth refactor', 'work on the next ticket', 'open the map for the payment effort', 'what's…

use-git-identity

Set the git identity (name + email) before any commit, amend, or rebase. Apply repo-local on first commit; use `-c` for one-offs.

forge-flow

Load when the user starts a new session on an ongoing effort — fresh chat, 'let's continue', 'pick up where we left off', or the first turn of any working session.

git-issue-tracker

Issue-tracker integration point for git-tracked repos. Premade with GitHub operations (gh CLI — native sub-issues for parent/child, native blocked_by dependencies for blocking, frontier queries). Load whenever a skill or workflow (e.g. wayfinder) needs to…

frontend-design

Load when building new UI or reshaping an existing one — visual design, typography, palette, layout decisions — or when a design reads as a templated default.

subagent-driven-development

Coordinator-driven parallel implementation — dispatch one implementer subagent per task in its own worktree, per-task review, fix loop, integrate. Load when the plan has independent tasks ready to execute.

ste100

Write human-facing text in ASD-STE100 Simplified Technical English — short sentences, approved words, one meaning per word. Governs PR bodies, docs, and issue/PR comments; chat replies and commit messages are not in scope (caveman-commit owns commits).

wayfinder

Plan a huge chunk of work — more than one agent session can hold — as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.

pr-review

Review a GitHub PR or local branch inside a worktree — hard-rule validation, then a kimi -p review loop producing caveman-review findings. PR mode posts ONE review under the authenticated user's identity; local mode returns findings. Triggers: "review this…

finishing-a-development-branch

Finish a development branch — commit, verify, push, open PR. PR is the only path. Load when implementation is complete and tests pass. Triggers: "finish this branch", "wrap up", "open a PR".

planning-and-task-breakdown

Breaks work into ordered tasks. Use when you have a spec or clear requirements and need to break work into implementable tasks. Use when a task feels too large to start, when you need to estimate scope, or when parallel work is possible.

pr-resolve

Resolve PR review findings — reads caveman-review findings, dispatches one resolver subagent per finding group into its own worktree, commits under user's git identity, pushes to PR head branch, replies under each review thread with the resolving commit SHA.…

loops

Drive `kimi -p` review/resolve loops via a shell framework. Dynamic prompts: callers supply their own `.md` template; loops renders it, compresses it, drives `run_loop` until `DONE:` or `BLOCKED:`. Triggers: "kimi -p loop", "review loop", "resolve loop".

forge-seo

Load when the task involves SEO — meta tags, sitemaps, structured data, rendering strategy, Core Web Vitals, search ranking, or content optimization.

forge-eu-accessibility

Load when building or auditing digital products for the EU market — accessibility work, WCAG compliance, accessibility statements — or when the user mentions BFSG, EAA, EN 301 549, or EU accessibility law.

forge-tailwindcss-conventions

Load when the task involves Tailwind CSS — styling components, utility classes, className work, converting custom CSS to utilities, or Tailwind build/config issues. Covers v4 (CSS-first) and flags v3 differences.

kiss-principle

Load when a design or implementation feels over-engineered, when evaluating architecture or reviewing code for simplicity, or when the user mentions KISS, over-engineering, or premature abstraction.

12-factor-app

Load when designing, building, or reviewing SaaS, cloud-native apps, microservices, containerized workloads, or serverless functions — or when the user mentions 12-factor, config in env, stateless processes, or dev/prod parity.

forge-docs

Load before code changes, before writing docs or an ADR, when a user question exposes a missing how-to, and when closing a chunk of work; also loaded by forge after each squash-merge and by deep-research after a report.

forge-init

Repo bootstrap runbook — writes the local agent contract and captures repo specifics. User-invoked only.

forge-setup

Harness adaptation runbook — clone the repo, find harness-specific references, patch them for the running harness. User-invoked only.

forge-cleanup

Load when the user wants to tidy up — 'clean up', 'remove stale stuff', after a merged PR, or when worktrees and scratch files pile up.

domain-modeling

Build and sharpen a project's domain model. Use when the user wants to pin down domain terminology or a ubiquitous language, record an architectural decision, or when another skill needs to maintain the domain model.

caveman-commit

Ultra-compressed commit message generator. Cuts noise from commit messages while preserving intent and reasoning. Conventional Commits format. Subject ≤50 chars, body only when "why" isn't obvious. Use when user says "write a commit", "commit message",…

caveman-review

Ultra-compressed code review comments. Cuts noise from PR feedback while preserving the actionable signal. Each comment is one line: location, problem, fix. Use when user says "review this PR", "code review", "review the diff", "/review", or invokes…

caveman

Ultra-compressed communication mode. Cuts output tokens 65% (measured) by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman…

conventional-commits

Format commits per Conventional Commits v1.0.0 spec. Load when writing any commit message.

deep-research

Exhaustive evidence-based research — minimum 10 search iterations, recursive reflection, structured reports with markdown-native tables. As long as the research is solid; no fixed length cap. Triggers: "research this", "deep dive", "exhaustive analysis",…

dispatching-parallel-agents

Dispatch up to 10 parallel subagents via AgentSwarm for 2+ independent tasks. Triggers: "swarm mode", "parallel agents", "dispatch in parallel", or 2+ independent tasks with no shared state.

grilling

Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, or uses any 'grill' trigger phrases.

prototype

Build a throwaway prototype to answer a design question. Use when the user wants to sanity-check whether a state model or logic feels right, or explore what a UI should look like.

resolving-merge-conflicts

Use when you need to resolve an in-progress git merge/rebase conflict. Loads `subagent-driven-development` when conflicts touch integration branches and need per-task unblock work.

using-git-worktrees

Create an isolated git worktree for feature work. Detect existing isolation first; only create if needed. Triggers: "worktree", "isolate", "parallel implementation", "subagent dispatch".

verification-before-completion

Verify completion claims with fresh evidence. Iron Law: NO COMPLETION CLAIMS WITHOUT FRESH VERIFICATION EVIDENCE. Load when about to claim complete/fixed/passing; before commit/PR/task done.