Scenario 31: Import dual-mode plugin repo
Use this when a repository contains both harness project files (for example AGENTS.md, CLAUDE.md) and a plugin manifest at the repo root (for example...
Frequency: Common · Status: Shipped
Use this when a repository contains both harness project files (for example
AGENTS.md, CLAUDE.md) and a plugin manifest at the repo root (for example
.claude-plugin/plugin.json, .codex-plugin/plugin.json, or
.github/plugin/plugin.json). A normal scan automatically merges
harness files with repo-root skills/, commands/, and hooks/ trees that the
plugin manifest describes.
Typical commands:
# Preview harness + plugin-source imports (merged when a manifest is present)
harnesstap scan . --dry-run
# Scope to one harness
harnesstap scan . --harness claude-code
# Build a layer from the merged scan (layer from-project always merges both)
harnesstap layer from-project my-layer --project .
harnesstap layer show my-layerDry-run output shows two sections when both sides contribute: harness platforms
and plugin imports. Imported plugin resources carry provenance metadata
(source_plugin_kind, relative_path) for audit and resource sync.
See portability limits for what plugin import does and does not cover (hooks with install-time paths, runtime-only adapters).
Scenario 28: One-command machine migration
Use this when moving HarnessTap state to a new machine or sharing your curated workspace with a teammate offline. The archive includes layers, environments...
Scenario 32: Apply to instruction-tier harnesses
Use this when applying a layer to harnesses that load skills as always-on instructions or rules rather than agent-requested skill directories. Harnesses...