ingest-brain

name

ingest-brain

description

Add new content to the user's local second-brain knowledge base using the `brain` CLI's ingestion commands. Use this skill whenever the user asks to ingest, save, pull, or capture content into the brain — including files (PDFs / DOCX / Markdown / TXT), directories of files, Gmail threads, Krisp meeting transcripts, Slack threads, or arbitrary stdin snippets. For Krisp/Slack the brain does NOT have a direct integration; instead orchestrate the relevant MCP server and pipe the fetched content into `brain ingest-stdin`. Always pick a sensible scope flag — never bulk-ingest the Gmail inbox. MANDATORY TRIGGERS: ingest, ingest last week, ingest this file, ingest this directory, pull this email, pull last week's Krisp calls, save this Slack thread, save this transcript, add this to my brain, capture this in my brain, import this into the brain, brain ingest, brain ingest-stdin, brain ingest-gmail.

Ingest Brain

Use the brain CLI to add content to the user's hybrid-search corpus. Pick the right ingestion path based on where the content lives. Never invent content — ingest only what the user has actually pointed you at or what an MCP returned.

For searching/synthesis, see consult-brain. For authoring brand-new vault notes, see brain-authoring.

Choosing the right command

Source	Command	Notes
One local file (PDF / DOCX / MD / TXT)	brain ingest <path>	Dedup by source_path. --force reapplies.
A directory of files	brain ingest-dir <dir>	Recursive. --ext md,pdf to scope. --dry-run first.
Gmail messages	brain ingest-gmail	REFUSES without a scope flag. See Gmail section.
Krisp transcript / Slack thread / arbitrary text	brain ingest-stdin	The MCP→stdin chain. See Krisp + Slack sections.
Krisp action items as a separate doc	brain enrich --krisp-action-items (handoff) + brain ingest-stdin --content-type krisp_action_items	See brain-todo for the full flow.

Files and directories

brain ingest ~/Downloads/2026-Q1-OKRs.pdf --tag okrs
brain ingest-dir ~/career-docs --tag career --dry-run    # preview first
brain ingest-dir ~/career-docs --tag career --ext md,pdf

Verbs in output:

• ingested: — new row inserted
• updated: — existing row replaced in-place (content or tags changed, or --force)
• skipped: — content unchanged

--force on brain ingest re-applies even when content is byte-identical. Useful after a chunker / embedder upgrade.

brain ingest-dir has no --force flag — it always UPDATEs in-place when the file changed and skips otherwise.

Gmail — brain ingest-gmail

This command refuses to run without a scope flag by design. Pick the narrowest scope that satisfies the user's ask:

User ask	Flag
"Ingest emails from <person>"	--from [e-mail]
"Ingest my Linear tickets / GitHub PR notifications"	--label "linear/notifications" or whatever the user uses
"Pull emails about "	--query "subject:<topic> OR body:<topic>" (Gmail query syntax)
"Pull last 30 days"	--since YYYY/MM/DD (concrete date 30 days ago — Gmail's native after: syntax; relative strings like 30d build a broken after:30d and return nothing)
"Pull this specific thread"	--query "rfc822msgid:<msg-id>" — --query takes raw Gmail search syntax only (e.g. from:[e-mail] subject:invoice). If the user gives you a Gmail search URL, extract the underlying query string from it before passing — don't paste the URL.

Always run with --dry-run first when the scope might be broader than the user expected. Cap with --max 100 for safety. Note: drafts are now included (Q1-A change) — they ingest with documents.draft=TRUE and are hidden from the wiki but visible to brain search / brain show.

The summary line is N ingested (M draft) when drafts are present.

Krisp transcripts — MCP→stdin chain

The brain has no direct Krisp integration. Claude orchestrates:

1. Call the Krisp MCP to search/list meetings, fetch transcripts. Use whichever Krisp MCP tool fits — your harness exposes them as mcp__claude_ai_Krisp__*. Don't hardcode parameter names; read your live MCP schema for the available tools and parameters.
2. For each meeting, pipe the transcript body into brain ingest-stdin:

echo "<transcript content>" | brain ingest-stdin \
  --source krisp \
  --external-id "<meeting_id>" \
  --title "<short title — usually the meeting topic + date>" \
  --content-type transcript \
  --date YYYY-MM-DD \
  --metadata '{"participants": ["Alice", "Bob"], "duration_min": 42}'

Dedup is by (source_kind='krisp', external_id=<meeting_id>) — re-ingesting the same meeting UPDATEs in place. UUID + tags survive.

3. Report counts back to the user: "Ingested N new transcripts, updated M existing ones."

Slack threads — MCP→stdin chain

Same pattern with the Slack MCP. Fetch a thread, then:

echo "<thread content>" | brain ingest-stdin \
  --source slack \
  --external-id "<channel-id>/<thread-ts>" \
  --title "Slack — <#channel> — <first message snippet>" \
  --content-type thread \
  --date YYYY-MM-DD \
  --metadata '{"channel": "#engineering", "participants": ["U1", "U2"]}'

Pick --external-id so it's stable across re-ingests. Channel-id + thread-ts is the canonical Slack identifier; thread-ts alone is also unique within a workspace.

Tags on ingest

Every ingest command takes one or more --tag flags. Tags are auto-normalized at write time (casefold-lowercase, hyphen-separated, deduped) — any case/separator at the CLI is fine.

brain ingest <file> --tag interview-prep --tag career-2026

Disabling the post-ingest summary

All four ingest commands run an LLM auto-summary post-hook by default (Q1-D). Skip with --no-enrich when:

• The doc is tiny and the summary would be noisy.
• Ollama is known down and you want the ingest to complete fast.
• You're scripting a bulk import and will run brain enrich --backfill afterwards.

brain ingest-stdin --source slack --external-id ... --title ... --no-enrich

Verifying after ingest

After any non-trivial ingest, sanity-check:

brain status                                   # totals + breakdown by source
brain list --source krisp --limit 5            # see the freshest rows
brain show <id-prefix>                         # verify content + summary

If brain doctor flags an Ollama / Postgres issue, route to brain-maintenance.

Safety rules

• Never bulk-ingest the inbox. brain ingest-gmail enforces this with the scope-flag requirement; respect the spirit when picking scopes.
• Don't invent content for stdin ingests. If the MCP returned nothing, tell the user — don't generate a placeholder transcript.
• No untagged dumps. Always pass --tag so the user can find/clean the batch later.
• Mind external_id stability. A Krisp meeting that gets re-ingested with a different --external-id becomes a duplicate row. Match what the MCP returns as the canonical meeting id.