Skill Claude CodeCodex
Use at the start of any non-trivial engineering task to load the right agent-framework doc(s) for the work — code review or PR prep, security review, pentest, hardening, threat modeling, supply-chain/dependency risk, debugging, quality gates, test-evidence ownership, Codex/Claude routing and delegation…
Cursor rule Cursor
Apply the repository Agent Configuration Framework before substantial coding-agent work.
Skill Codex
Route bounded and bulk work to Codex, keep architecture, security, authentication, data-loss, release, and final-review work on Claude, and run Hermes PR reviews through Codex (model set in Hermes's own config). Use when selecting a model lane, coordinating a bounded delegate, or deciding effort and escalation.
Command Claude Code
Run a bounded cmux + Hermes orchestration step via the local deterministic broker — token-free doctor, write-isolated lane/worktree creation, fail-closed noninteractive advisor, persistent Hermes master in remote tmux, explicit UUID send, exact usage, and preserving cancel/close. No reverse SSH, environment…
Command Claude Code
Arbitrate between competing plans or execution lanes on evidence, capability, budget, and risk; pick the efficient-frontier handoff, partition writes by disjoint ownership, watchdog-verify delegated results independently, and enforce stay-with-limits budget checks before each bounded wave. Builder.io concepts under…
Skill Codex
Orchestrate bounded, default-off work between cmux (local UI/session transport and lifecycle) and Hermes (provider router, plan/delegation brain, fallback, and usage ledger on a Tailscale-only VPS) using a local deterministic broker. Use when an operator has adopted cmux as the local surface and Hermes as the remote…
Skill Codex
Arbitrate between competing plans or execution lanes on evidence, capability, budget, and risk instead of first-past-the-post. Use when two or more plans, agents, or model lanes could plausibly own a task and the coordinator must pick the efficient-frontier handoff, watchdog-verify delegated results, and enforce…
Skill Codex
Gate and use optional repository context accelerators such as Graphify-compatible graphs/context maps, OpenWiki-compatible generated agent wikis, symbol indexes, or code-review graphs. Use when the user or repo opts into one of these tools, asks for token-saving context strategy, scope mapping, blast-radius analysis…
Skill Codex
Route into the albrand/ai-config-kit operating framework and load only the relevant doctrine for planning, implementation, debugging, review, security, harness design, quality convergence, repository adoption, journaling, skill learning, token economy, or context acceleration. Use when the user asks to apply, inspect…
Command Claude Code
Read an existing project deeply, generate knowledge docs, identify missing requirements and tickets, and plan hardening work against source evidence.
Command Claude Code
Bootstrap or reconcile a project roadmap, operating model, and ticket ecosystem from business docs, designs, legacy boards, and stakeholder prompts.
Command Claude Code
Bootstrap or reconcile the technical platform, local environments, CI/CD gates, PR automation, infrastructure, and AI developer workstreams.
Skill Claude CodeCodex
Perform a whole-project assessment, generate or refresh knowledge docs, derive missing requirements and tickets from internal and external sources, and plan hardening work. Use when the user wants an existing project read deeply and turned into docs, gaps, QA logic, and implementation plans.
Skill Claude CodeCodex
Turn business requirements, docs, designs, existing boards, legacy roadmap material, and stakeholder prompts into an evidence-backed roadmap, operating model, and PR-sized ticket ecosystem. Use when the user wants AI to bootstrap, import, reconcile, or create product/project roadmap artifacts for a new or legacy app.
Skill Claude CodeCodex
Bootstrap or reconcile the technology side of a project from a stack prompt, existing repos, infrastructure choices, cloud/provider access, local environment needs, CI/CD gates, PR automation, and AI developer workstreams. Use when the user wants a project-agnostic technical framework setup plan or live setup actions.
Command Claude Code
Turn a module description into an evidence-backed delivery plan with phases, PR-sized tickets, resources, risks, and validation.
Skill Claude CodeCodex
Turn a product or platform module idea into an evidence-backed delivery plan using AI analysis plus normal repository, documentation, and board inspection. Use when a user provides a module name, capability area, roadmap item, Linear/project-planning request, or migration target and wants Codex or Claude Code to study…
Skill Codex
Discover and use the host software an agent runs in or through — local session transports (cmux), terminal multiplexers (tmux, zellij), and other agentic shells/harnesses — via a capability-first, host-neutral adapter contract. Use when an agent needs to detect which native surfaces are present, route on declared…
Skill Codex
Enforce isolated Orca-only interactive browsing and ownership-safe browser page hygiene. Use for website navigation, browser E2E, visual inspection, authentication, screenshots, DOM, console, network, or page interaction.
Skill Codex
Explicit-only skill for Orca-owned workflow automation, including per-repository PR listeners, scheduled review-queue sweeps, execution-productivity telemetry, and optional read-only Hermes critique. Orca owns schedules, worktree/workspace lifecycle, terminal targeting, and browser/mobile/emulator control. Review work…
Command Claude Code
Review a pull request or diff with business-rule-first analysis, high-signal filtering, scoped instructions, validated findings, mandatory output-contract loading, inline code threads, root-cause commentary, practical failure examples, GitHub suggestion blocks when safe, no monolithic review bodies, no…
Skill Codex
Review pull requests or diffs with a business-rule-first, high-signal, low-false-positive workflow, mandatory PR review output contract loading, inline code threads, root-cause commentary, practical failure examples, GitHub suggestion blocks when safe, no monolithic review bodies, no validation-transcript boilerplate…
Command Claude Code
Multi-pass, multi-lens security review that reinforces detection through independent blind finder passes and an adversarial refute pass, then reports only confirmed findings. Static by default; owned or authorized targets only.
Command Claude Code
Authorization-gated white-hat penetration test of an owned or explicitly authorized target, following find → validate → fix → regress. Requires an authorization gate before any active testing.
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: