PDF ingestion linking note retains misleading pending text and lacks useful navigation #1651

During the Memory Quest PDF walkthrough, the generated ingestion-run (linking) note continued to say:

Raw extraction accepted; exact storage materialization is pending.

Extraction had succeeded, the content note was readable, and the materialization jobs had run. The message looks like a current status and incorrectly suggests processing is still incomplete.

Code evidence

src/basic_memory/document_ingestion/raw_document.py, build_raw_ingestion_run_markdown, hardcodes this sentence into the run note body. It is creation-time text, not a live status indicator. The note's frontmatter already carries source and output references.

Proposed change

• Replace the transient pending message with permanent, user-facing wording explaining that this note records the PDF extraction and connects the original document to its extracted text.
• Include clickable links to the original PDF and the extracted content note, using supported note/file link conventions.
• Avoid claiming that storage work is complete merely because extraction was accepted. If live processing status is needed, derive it from actual state rather than static prose.
• Address existing generated notes containing the exact placeholder. Preserve user edits and extraction provenance; define a safe repair/backfill strategy rather than overwriting arbitrary note bodies.

Acceptance criteria

• Newly generated linking notes explain their purpose and let the user open both the source PDF and extracted content.
• Successful extraction does not leave a misleading permanent pending message.
• Existing unedited placeholder notes have a supported repair path.
• Tests cover generated text, link targets, and protection of user-edited content during any repair.

Discovered while testing the cloud Memory Quest PDF path. Keep this separate from the Memory Quest UI worktree; the generator is in core.