Show the evidence matrix

<!-- sflow-copilot-pause -->

Before any boundary lookup or SFlow action, run singularity-flow pause status --json and follow its data.agentInstruction. If data.paused, answer as native Copilot, only offer /sf-pause off and never resume implicitly.

<!-- sflow-output-contract: concise-relay -->

Output contract: Relay requested CLI fields or output faithfully; preserve warnings/errors and only the explanations required below. Summarize artifacts with verified Markdown-preview links; never paste document bodies, code or diffs into Copilot. For code, list touched paths and observed added/deleted/created line counts; unavailable stays unavailable. Use existing evidence, not extra calls solely for display. Summaries, links and opened previews are not human review or approval. For every next action, honor returned commandGuidance in order: Copilot, Shell, VS Code (model-free) using modelFreeCommand; if absent, say "not available for this action". Missing Copilot: "no verified equivalent". Never invent routes.

<!-- sflow-execution-boundary -->

Boundary: no Story required; cwd=opened Git root or verified repositoryPath from singularity-flow workspace current --json; refuse if neither resolves; never search $HOME/parents. Use exact CLI-returned paths. repositoryPath is command cwd, not a prefix for absolute file paths. Never rebase paths or strip .git/worktrees. Issued private/external files are exact exceptions to repository confinement, not wider read/write permission; writes still need authorization. Resolve relative paths only against their declared base. On failure, preserve the path and error; never search or recreate it.

1. Use the Boundary repository. With an explicit Work ID, do not require or change the active Story. Without an ID, run singularity-flow session current --json and require its ready Story; use its exact repositoryPath as cwd.
2. Run singularity-flow evidence matrix <WORK-ID> --json, or without the ID for the attached Story, adding only --row, --result or --facet when the person asked for them.
3. From data.matrix.evaluation, show completion.label with its reasons, lifecycle.words, the counts in summary.results, summary.assuranceFloor, and summary.scope.words and summary.scopeRevision.words when present. Then show page.rows as a table: row ID, result, assurance, and each obligation's responsibility and status.
4. Use only the labels the CLI returned. Never say a Story is complete, all tests passed or all requirements are satisfied unless completion.label says so; "module-observed" means a covering test command passed; only "exact-local-observed" means the criterion's own test was seen passing.
5. List findings that need attention and each row's actions commands as next steps, without running them.
6. When the person asks which requirement statements the Story's sources hold, run singularity-flow evidence scope <WORK-ID> --json instead: show each item's id, text, sourceId and disposition, the unreadable sources, structurallyComplete and completenessReview. A review never means the scope is correct.
7. Do not change files or lifecycle state.