Skill Claude Code
GitHub Copilot code review using thermo-nuclear correctness and code-quality rubrics. Use for pull request review, branch diffs, and review of added or modified code.
Skill Claude CodeCodex
Shared vocabulary for designing deep modules. Use when the user wants to design or improve a module's interface, find deepening opportunities, decide where a seam goes, make code more testable or AI-navigable, or when another skill needs the deep-module vocabulary.
Skill Claude Code
Diagnosis loop for hard bugs and performance regressions. Use when the user says "diagnose"/"debug this", or reports something broken/throwing/failing/slow.
Skill Claude Code
Sync issue-tracker tickets into docs/ markdown files. Use when the user wants to export or refresh project backlog docs from an issue tracker.
Skill Claude Code
Grill the user relentlessly about a plan, decision, or idea. Use when the user wants to stress-test their thinking, uses any 'grill' trigger phrases, or when another skill needs the grill.
Skill Claude Code
Grounded research. Traceability-first research workflow for evidence, reliability, product, policy, technical, sentiment, complaint, review, public-opinion, and source-backed questions. Use when the user asks for grounded research, rigorous research, exact source traceability, text fragment links, citations…
Skill Claude Code
Evidence discipline for factual claims. Receipts before claims, no guessing, unknown is an allowed answer. Use as the baseline standard when citing sources, files, or command output, and when the user asks you to re-check an answer you already gave because they cannot verify it themselves.
Skill Claude Code
Compact the current conversation into a handoff document for another agent to pick up.
Skill Claude Code
Drive a development task — any size — through the developer-led loop (capture intent, plan, execute, reconcile), keeping the developer in control and oriented. Type /helm to start or resume.
Skill Claude CodeCodex
Scan a codebase for deepening opportunities, present them as a visual HTML report, then grill through whichever one you pick.
Skill Claude Code
use when the user wants a simple worded reply, or expresses that your speech is too complex, jargon-heavy, hard to understand etc. Or when they say "keep it simple".
Skill Claude Code
Answer questions about a project's context, issues, and domain by searching docs. Use when the user asks any question about project-level knowledge scoped to a named project folder under docs/.
Skill Claude Code
Report the delta between a planned change and what was actually built — deviations, unplanned decisions, whether the result meets the intent, and any docs to update. Use to close out a task before calling it done, whenever a plan or spec was followed; helm's Reconcile phase reaches for this. Especially important after…
Skill Claude Code
Install the /agentic-loop delegated build loop into this repo — orchestrator, implementer, and an independent reviewer that runs in its own context — fitted to the repo's existing layout.
Skill Claude Code
Delegated build loop — approve a spec and its slices once, then implementation, verification, and independent review run without you. Type /agentic-loop .
Agent Claude Code
Independent read-only judge for slices that change structure — new modules, cross-boundary dependencies, new interfaces, or added packages — returning a one-line verdict.
Agent Claude Code
Implements one slice against its numbered spec clauses, and on a fix round reads the verify log or reviewer findings it was handed and clears what failed.
Agent Claude Code
Runs an approved spec slice by slice — delegates implementation, runs the verify command, dispatches independent review, bounds the fix rounds, and reports once.
Agent Claude Code
Independent read-only judge for one slice — renders a verdict per spec clause with receipts, writes the findings file, and returns a one-line verdict.
Agent Claude Code
Independent read-only judge for slices that touch the trust boundary — untrusted input, authentication, authorization, secrets, or dependencies — returning a one-line verdict.
Agent Claude Code
Independent read-only judge for slices touching {{SURFACE}} ({{GLOBS}}) — reviews for {{FOCUS}} and returns a one-line verdict.
Skill Claude CodeCodex
Project an existing canonical .claude setup to GitHub Copilot, or fully AI-enable a repository for Claude Code and Copilot with shared instructions and the developer-led workflow. Re-runnable; converges the repo on the layout.
Skill Claude Code
Test-driven development. Use when the user wants to build features or fix bugs test-first, mentions "red-green-refactor", or wants integration tests.
Skill Claude Code
Synthesize the settled conversation into a thin Plan on the work-tree leaf doc — no interview. Use when intent is pinned and you need a durable plan before building; helm's Plan phase reaches for this when the goal fits in one change.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: