Current operating surface

Use this as an active operating surface or current narrative. Keep this as one of the main working surfaces and use it to direct people to the next action.

← Reports
Grant Frontier

Grant Source Control Surface

This is the live operating view for how new grants get into CivicGraph: government feeds, foundation pages, frontier queues, and the automation rail that decides whether the system is growing or just sitting on latent URLs.

How to use this

This page is not where ACT decides what to apply for. It is the source-control room for the grant machine: check whether the database is fresh, which source URLs need scouting next, which crawlers are failing, and which high-value funders still have no structured program data.

1. Check pressure
Use due frontier checks to see whether the scout backlog is under control.
2. Fix failures
Use blocked frontier rows to repair URLs, parser hints, or agents.
3. Feed decisions
Use uncovered funders and new program hits to improve Matched, Tracker, Home, and the ACT cockpit.
0
Grant rows
0
Future deadlines
0
Frontier URLs
0
Due frontier checks
Pipeline funnel

Raw discovery → verified → ACT recommendations → decisions

One vertical view of how this session's work flows: scrapers produce raw rows, promotion + LLM auto-classify lifts them into alma, the recommendations MV scores against 6 ACT projects, the user's decisions feed the blocklist and Notion.

Funder context layer (relationship temperature for the recommendations dossier)
0
funders with context snapshot
0
WARM (≥50 score · Xero/contacts)
0
TEPID (20–49 · partial signal)
0
LIGHT/COLD (build-before-ask)
What this tells us

Where the grant machine is strong and where it is barely switched on

Government feed base
0

rows currently marked open-like. The grant table already has scale; the main gap is freshness and breadth, not raw existence.

Foundation linkage
0

grant rows already linked to a foundation, with 0 open structured foundation programs sitting behind 0 foundation websites.

Untouched frontier
0

frontier rows have never succeeded yet. This is the clearest answer to “what haven’t we tried?” — most of the frontier exists, but has barely been harvested.

Underworked queues

What we still have barely touched

Foundation base

How much philanthropic surface we can still convert

0 foundations
0 with giving data
0 with websites
the raw homepage pool available for frontier harvesting
0 structured programs
0 currently marked open
Automation rail

Which grant discovery agents are actually live

These are the rails that decide whether the grants system keeps expanding or just sits on yesterday’s rows.

Source movement

First grant frontier baseline

This is the beginning of the same control-loop we built for youth justice: source families that are changing, getting blocked, or sitting cold despite being high-value.

Blocked frontier

Which specific URLs are failing right now

No failing frontier URLs are currently recorded.
Current intake mix

Where grant rows are actually coming from

Frontier pressure

What the crawler should work next

High-value uncovered funders

Which long-tail funders still have zero structured programs

This is the cleanest national expansion queue right now: funders with websites, money, and likely program surfaces, but no structured program rows yet. The frontier now seeds extra candidate paths for this set.

Long-tail discovery hits

What the new long-tail rail has already turned into structured programs

No structured programs from the long-tail rail yet. The discovery agent is now targeted at this queue; this panel will start to fill as those runs land.