---
name: oso-filecoin-pgf
description: Analyze Filecoin Public Goods Funding (PGF) and ecosystem health via the Open Source Observer (OSO) data warehouse — funding ROI across ProPGF / RetroPGF / Impact Grants, developer activity, onchain impact (data onboarded, SPs, block rewards, Filecoin Pay, warm storage / PDP), Karma milestones vs actual metrics, dependency & downstream impact, and network health. Load whenever the user asks about Filecoin project funding, ROI, developer metrics, milestone progress, dependency / downstream impact, onramp attribution, or wants to query the `filecoin.filpgf_public.*` mart layer (Trino SQL via OSO Chat, pyoso, or oso.xyz/filecoin). Complements the `filecoin-data-portal` skill: that one is raw parquet network data; this one is project- and program-level analytics built on top. Body simply points to the canonical OSO SKILL.md.
---

# OSO Filecoin PGF & Ecosystem Data

For the canonical, always-up-to-date instructions on how to query the Open Source Observer (OSO) data warehouse for Filecoin Public Goods Funding analysis — including the schema DAG (`filecoin.data_portal.*` → `staging_*` → `entities` → `events` → `metrics` → `filpgf_public` mart layer), the Trino SQL dialect notes, the metric catalog discovery query, the Karma-milestones-to-actual-metrics workflow, the full table inventory, and the canonical starter queries — load:

**https://raw.githubusercontent.com/opensource-observer/insights/refs/heads/main/community/filecoin/skills.md**

That document is maintained by the OSO team in the [opensource-observer/insights](https://github.com/opensource-observer/insights) repo and is the single source of truth. Treat its connection instructions, schema overview, and SQL examples as authoritative and follow them directly.

## Why this skill is just a pointer

The upstream `skills.md` is the file OSO ships to its own agents and notebooks. Tables move, metrics get added, schemas evolve as new ingestion sources land (Karma, DataCapStats, Drips, Filecoin Pay, warm storage, dependency survey, …). Duplicating it here would only create drift. Fetch the URL, follow it, and you have the freshest schema and workflow every time.

## When to load this vs the Filecoin Data Portal skill

- **`filecoin-data-portal`** — raw parquet network data, queryable with DuckDB. Use for direct network telemetry: storage providers, clients, deals, PDP, Filecoin Pay rails, daily metrics.
- **`oso-filecoin-pgf`** (this skill) — project- and program-level analytics built on top of that data. Use for funding ROI, developer activity, milestone tracking, dependency impact, and cross-program comparisons.

The two are complementary; loading both is appropriate when a question spans network-level facts and project-level attribution.

## Quick links

- Canonical SKILL.md: https://raw.githubusercontent.com/opensource-observer/insights/refs/heads/main/community/filecoin/skills.md
- Live dashboard: https://www.oso.xyz/filecoin/Public-Goods-Funding
- Public repo: https://github.com/opensource-observer/insights
- Dependency survey: https://dependency-survey.pages.dev
- OSS Directory (project slug source of truth): https://github.com/opensource-observer/oss-directory
- API key signup: https://www.oso.xyz/start