harnessloop-channelslisted

# Harnessloop Channels List all declared external systems, channels, and tools that Harnessloop may rely on. This skill produces an inventory; it does not test connectivity, infer missing tools, or call external systems. ## Input Contract Accept an explicit skill invocation such as `$harnessloop-channels`, or natural language asking to list external systems/tools/channels. Treat `harnessloop:channels` as a natural-language alias only; `$harnessloop:channels` is not a valid skill invocation. Useful input includes: - `target-project`: defaults to the current working directory. - Optional scope: setup, evidence, active goal, active round, intake, handoffs, or all. - Optional output preference: table, grouped by system, grouped by tool, or missing-fields list. If `.harnessloop/` is missing, report `not-initialized` and suggest `$harnessloop-init`. ## Processing Contract 1. Read declared sources from `.harnessloop/setup/data-sources.md`, `.harnessloop/state/evidence-index.md`, active goal `data-contract.md`, active round evidence/review files, intake transfer packets, and open handoffs when present. 2. Extract only declared external systems, tools, channels, accounts, credential references, local parameter references, permissions, required parameters, verification methods, failure handling, and sensitivity labels. 3. Mark missing or ambiguous fields explicitly instead of inferring them. 4. Classify each channel as `declared-complete`, `declared-incomplete`, `unused`, `unkn