v01/SKILL.md
Use when the user explicitly invokes $vibeloom or asks to initialize, import, review, fix, reconcile, or evolve a governed codebase through VibeLoom's strict command interface, surface modes, and contract stack of intent, prd, usm, dm, and spec.
npx skillsauth add ilya-baimetov/vibeloom vibeloomInstall 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.
Use this skill only when the user explicitly invokes $vibeloom or clearly asks to operate through the VibeLoom workflow.
This skill is for governed, contract-driven work over the canonical stack:
intent.mdprd.mdusm.mddm.mdspec.mdSKILL.md is the runtime entrypoint and orchestrator. The canonical layer contract lives in docs/vibeloom-methodology.md; routine runtime authority lives in references/.
On each invocation, start with the smallest correct reference set:
Do not load docs/ during routine commands unless the user asks for help, requests deeper explanation, or a runtime reference explicitly escalates.
help is the only command family allowed to load explanatory material outside references/.
Topic routing:
methodology -> docs/vibeloom-methodology.mdprofiles -> docs/profile-selection.mdsurfaces -> docs/surface-modes.mdevals -> docs/evals-structural.md, docs/evals-semantic.mdtemplates -> templates/commands -> references/command-surface.mdRules:
references/ may escalate only by topic name, never by direct docs/* path.help <topic> as the canonical explanation entrypoint.references/ plus required canonical artifacts and templates unless help is invoked.Additional docs:
docs/artifact-protocol.md and docs/context-loading.md are maintainer-facing supporting docs.help topics and should not be loaded by routine runtime references.$vibeloom with no /vibeloom ... command uses state-aware triage from references/routing-and-loading.md./vibeloom help methodology and /vibeloom help templates.SKILL.md.AGENTS.md or plan.md as canonical semantic authority.USM or DM from the governed methodology.When updating this skill or its runtime references, validate behavior with references/acceptance-tests.md.
development
Develops vibeloom itself. Adversarial consistency checks and generation across intent / canon / skill / site for any vibeloom version. Use to bring up a new version, eval/review canon for drift, generate downstream artifacts from upstream specs, or run cross-agent reviews across multiple agents.
development
--- name: vibeloom description: Contract-driven governance for long-lived AI-coded projects. Use when the user wants to bootstrap, import, generate, review, reconcile, or approve artifacts in a project governed by the VibeLoom methodology (modes: vibe, pm, dev, expert). argument-hint: "[init|import|generate|eval|review|reconcile|approve|status] [target]" --- # VibeLoom VibeLoom governs long-lived AI-coded projects by generating a tiered contract of specifications (`intent-specs` → `product-spe
development
--- name: vibeloom description: Contract-driven agentic engineering for long-lived AI-coded projects. Use when the user wants to bootstrap, import, generate, eval, review, reconcile, or approve artifacts in a project governed by VibeLoom (modes: vibe, pm, dev, ux, expert). argument-hint: "[init|import|generate|eval|review|reconcile|approve|status] [target]" --- # VibeLoom VibeLoom is the reference instantiation of the **codæ** paradigm (contract-driven agentic engineering). It governs long-liv
development
Maintainer-only workflow for handling GitHub Secret Scanning alerts on OpenClaw. Use when Codex needs to triage, redact, clean up, and resolve secret leakage found in issue comments, issue bodies, PR comments, or other GitHub content.