Single Aming Claw skill entrypoint. Use this first for Aming Claw governance, backlog, graph, runtime, worker, QA, dashboard, MCP, install/start/bootstrap, or dogfood work; prefer the MCP onboard_route_guide tool before choosing a role or successor contract.
Public demo for Backlog Duplicate: a new requirement overlaps existing backlog work, and the observer asks whether to merge, supersede, or keep it separate. Use when a user asks to run, preview, or collect duplicate backlog evidence with the current Claude Code or Codex observer session.
Public demo for Docs Drift: a feature changes but documentation is stale, and the drift state is surfaced before a second doc fix. Use when a user asks to run, preview, or collect evidence for docs drift with the current Claude Code or Codex observer session.
Public HN challenge entrypoint for Aming Claw. Use when a user asks to run, preview, present, or collect evidence for the multi-agent challenge: one observer coordinates multiple commit-bound workers, records a failed or interrupted worker, replays it from the same contract lineage, reconciles the target graph, and writes an audit self-review.
HN demo case for the fear that code changes leave docs, tests, and config stale after implementation. Guides evidence collection for Asset Inbox, binding state, Baseline and Possible drift, Review Queue, impact scope, and review boundaries.
HN demo case for the fear that AI does not understand project structure and duplicates work. Guides evidence collection for graph-first discovery, backlog contract, target file fence, and acceptance criteria before implementation.
HN demo case for the fear that AI and subagents are black boxes during implementation. Guides evidence collection for timeline DAG, observer and subagent lanes, dispatch gate, evidence inspector, isolated worktrees, and append-only contract evidence.
Guided operator for the Aming Claw HN multi-agent challenge demo. Use when a user asks to run, preview, present, or collect evidence for the HN demo showing one observer coordinating multiple commit-bound workers, failed/interrupted worker replay, graph traces, reconcile, and audit self-review.