feat: summarize observation review timelines

[cafitac]

Summary

• Add per-memory-ref observation windows to observations audit top refs.
• Add top-level observation_count / candidate_count and fact status-history summaries to observations review-candidates.
• Update dogfood docs and handoff so historical injections are easier to distinguish from current lifecycle state.

Test Plan

• TMPDIR=$PWD/.tmp-test uv run pytest tests/test_cli.py::test_python_module_cli_observations_audit_reports_frequent_and_stale_refs_without_raw_queries tests/test_cli.py::test_python_module_cli_observations_review_candidates_explains_top_refs_without_mutation_or_raw_queries tests/test_cli.py::test_python_module_cli_observations_audit_reports_low_signal_empty_retrievals -q
• TMPDIR=$PWD/.tmp-test uv run pytest tests/ -q
• uv run python scripts/check_release_metadata.py
• uv run python scripts/smoke_release_readiness.py
• npm pack --dry-run
• git diff --check
• node --check bin/agent-memory.js
• local DB smoke: PYTHONPATH=src uv run python -m agent_memory.api.cli observations review-candidates /Users/reddit/.agent-memory/memory.db --limit 200 --top 10 --frequent-threshold 3
• static diff secret scan: finding_count=0