feat: discover contributed repos via GitHub GraphQL API

The REST /user/repos endpoint only returns repos where the user is
owner, collaborator, or org member. Repos contributed to via PRs
(e.g. polkadot-js/api, zed-industries/zed) were never discovered
and their commits were missing from moments.

Now supplements /user/repos with a GraphQL
repositoriesContributedTo query, which returns all repos the user
has committed to, opened issues/PRs on, or reviewed — with cursor-
based pagination and no result cap.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

.gitignore

@@ -2,6 +2,7 @@
 **/*.rs.bk
 .env
 .env.local
+.zed/
 
 # frontend
 /ui/node_modules

CLAUDE.md

@@ -0,0 +1,78 @@
+# CLAUDE.md
+
+This file provides guidance to Claude Code (claude.ai/code) when working with code in this repository.
+
+## Project Overview
+
+**moments** is a personal activity timeline and portfolio site. It ingests developer activity from multiple forges (GitHub, Gitea, Mercurial, Bugzilla), stores raw JSON payloads in PostgreSQL, and serves a React frontend showing contribution graphs, a ranked project dashboard, and a filterable activity timeline.
+
+## Architecture
+
+Hexagonal (ports & adapters) Rust backend with a React/TypeScript frontend.
+
+### Crate Dependency Graph
+
+```
+moments-entities   — pure types/DTOs, no DB or HTTP deps
+       ^
+moments-core       — port traits (EventReader, EventWriter, EventSource, PollerStateStore)
+                     + presentation reshape + poller loop
+       ^
+moments-data       — sole adapter: PgStore implements all core traits
+                     + EventSource impls (github, gitea, hg, bugzilla)
+                     + SQL migrations
+       ^
+moments-api        — axum HTTP API binary (read-only, connects as moments_ro)
+moments-worker     — ingestion daemon binary (runs migrations, connects as moments_rw)
+```
+
+### Key Design Decisions
+
+- **Raw payload storage**: upstream JSON is stored verbatim in `events.payload` (JSONB). The `reshape()` function in `moments-core/src/presentation.rs` transforms payloads into `TimelineItem` at request time — no re-ingestion needed to change presentation.
+- **Public/private gate**: `events.public` boolean controls API visibility. Only `public = true` rows are served.
+- **Wire types are hand-maintained**: `ui/src/api/client.ts` mirrors Rust entity types manually.
+- **Migrations**: run automatically on worker startup via `sqlx::migrate!`. The API binary never runs migrations.
+
+### Frontend
+
+React 19 + Vite 6 (SWC) + TypeScript + Bootstrap 5. State/data via `@tanstack/react-query`. Package manager is **pnpm**.
+
+Routes: `/` (dashboard), `/activity` (timeline), `/project/:source/*` (project detail), `/cv` (resume).
+
+## Build & Dev Commands
+
+### Rust
+
+```sh
+cargo build --workspace              # build all crates
+cargo build --workspace --release    # release build
+cargo clippy --workspace             # lint
+cargo fmt --check                    # format check
+cargo test --workspace               # run tests
+
+# Run binaries (need DATABASE_URL)
+DATABASE_URL=postgres://localhost/moments cargo run -p moments-api
+DATABASE_URL=postgres://localhost/moments cargo run -p moments-worker
+```
+
+### Frontend
+
+```sh
+cd ui
+pnpm install                         # install deps
+pnpm dev                             # dev server on :5173 (proxies /api/* to localhost:8080)
+pnpm lint                            # tsc --noEmit type-check
+pnpm build                           # production build (tsc -b && vite build)
+```
+
+## Database
+
+PostgreSQL with three migrations in `crates/moments-data/migrations/`. Two roles: `moments_rw` (worker, full access) and `moments_ro` (API, SELECT-only).
+
+## API Endpoints
+
+All under `/v1/`: `healthz`, `events`, `sources`, `projects`, `activity/daily`, `forge/{source}/*`, `og/contributions.png`.
+
+## Deployment
+
+Production uses `./script/deploy.sh`. Services run under systemd with hardened units. Secrets resolved from `pass` store via template substitution. Nginx reverse-proxies `/api/` to the API host.

crates/moments-api/src/main.rs

@@ -170,7 +170,7 @@ async fn og_contributions(
         .iter()
         .filter_map(|s| s.earliest)
         .min()
-        .unwrap_or_else(|| Utc::now())
+        .unwrap_or_else(Utc::now)
         .date_naive();
     let today = Utc::now().date_naive();
 

crates/moments-data/src/github_repo.rs

@@ -1,16 +1,20 @@
 //! Per-repo commit enumeration for full GitHub history.
 //!
-//! The Search API caps at 1000 results; this source enumerates all repos
+//! Discovers repos via two sources:
-//! the user can access via `/user/repos` and walks each repo's commit
+//!   1. REST `/user/repos` — repos where the user is owner, collaborator,
-//! history via `/repos/{owner}/{repo}/commits?author={user}` — no cap.
+//!      or org member.
+//!   2. GraphQL `repositoriesContributedTo` — repos the user has committed
+//!      to, opened issues/PRs on, or reviewed, even without collaborator
+//!      status. No result cap (cursor-paginated).
+//!
+//! Then walks each repo's commit history via
+//! `/repos/{owner}/{repo}/commits?author={user}` with a `since` cursor
+//! to avoid re-fetching known commits.
 //!
 //! Events use `github-commit:{sha}` as their ID, matching the scheme in
 //! `github_search`, so duplicates are resolved via idempotent upsert.
-//!
-//! Per-repo poller state keys (`github-repo:{owner}/{repo}`) track which
-//! repos have been fully backfilled. First run paginates the full history;
-//! subsequent runs fetch only page 1.
 
+use std::collections::HashSet;
 use std::sync::Arc;
 
 use async_trait::async_trait;
@@ -114,22 +118,132 @@ impl GithubRepoSource {
                 break;
             }
         }
+
+        // Supplement with repos from GraphQL repositoriesContributedTo.
+        // This catches repos where the user contributed via PRs but isn't
+        // an owner, collaborator, or org member — no result cap.
+        let mut known: HashSet<String> = repos.iter().map(|r| r.full_name.clone()).collect();
+        let contributed = self.discover_contributed_repos().await;
+        match contributed {
+            Ok(extra) => {
+                for r in extra {
+                    if known.insert(r.full_name.clone()) {
+                        repos.push(r);
+                    }
+                }
+            }
+            Err(e) => {
+                warn!(error = %e, "GraphQL contributed-repos discovery failed; continuing with known repos");
+            }
+        }
+
         Ok(repos)
     }
 
-    /// Fetch commits for a single repo, paginating fully on first run.
+    /// Discover repos the user has contributed to via GraphQL.
+    /// Uses cursor-based pagination with no result cap.
+    async fn discover_contributed_repos(&self) -> Result<Vec<Repo>, SourceError> {
+        let token = match &self.config.token {
+            Some(t) => t,
+            None => return Ok(vec![]),
+        };
+
+        let mut repos = Vec::new();
+        let mut cursor: Option<String> = None;
+
+        loop {
+            let after = match &cursor {
+                Some(c) => format!(", after: \"{}\"", c),
+                None => String::new(),
+            };
+            let query = format!(
+                r#"{{ user(login: "{}") {{ repositoriesContributedTo(first: 100, contributionTypes: [COMMIT, PULL_REQUEST, ISSUE]{}) {{ pageInfo {{ hasNextPage endCursor }} nodes {{ nameWithOwner isPrivate }} }} }} }}"#,
+                self.config.user, after
+            );
+            let body = serde_json::json!({ "query": query });
+
+            let resp = self
+                .client
+                .post("https://api.github.com/graphql")
+                .header(header::AUTHORIZATION, format!("Bearer {token}"))
+                .header(header::USER_AGENT, USER_AGENT)
+                .header(header::CONTENT_TYPE, "application/json")
+                .json(&body)
+                .send()
+                .await
+                .map_err(|e| SourceError::Http(e.to_string()))?;
+
+            if !resp.status().is_success() {
+                return Err(SourceError::Http(format!(
+                    "{} POST graphql",
+                    resp.status()
+                )));
+            }
+
+            let data: Value = resp
+                .json()
+                .await
+                .map_err(|e| SourceError::Parse(e.to_string()))?;
+
+            // Check for GraphQL-level errors
+            if let Some(errors) = data.get("errors").and_then(Value::as_array) {
+                if let Some(msg) = errors.first().and_then(|e| e.get("message")).and_then(Value::as_str) {
+                    return Err(SourceError::Http(format!("GraphQL error: {msg}")));
+                }
+            }
+
+            let contributed = &data["data"]["user"]["repositoriesContributedTo"];
+            let nodes = contributed["nodes"].as_array();
+            if let Some(nodes) = nodes {
+                for node in nodes {
+                    let full_name = node
+                        .get("nameWithOwner")
+                        .and_then(Value::as_str);
+                    let private = node
+                        .get("isPrivate")
+                        .and_then(Value::as_bool)
+                        .unwrap_or(false);
+                    if let Some(name) = full_name {
+                        repos.push(Repo {
+                            full_name: name.to_string(),
+                            private,
+                        });
+                    }
+                }
+            }
+
+            let has_next = contributed["pageInfo"]["hasNextPage"]
+                .as_bool()
+                .unwrap_or(false);
+            if !has_next {
+                break;
+            }
+            cursor = contributed["pageInfo"]["endCursor"]
+                .as_str()
+                .map(String::from);
+        }
+
+        debug!(repos = repos.len(), "discovered contributed repos via GraphQL");
+        Ok(repos)
+    }
+
+    /// Fetch commits for a single repo, paginating fully on first run
+    /// and using `since` on subsequent runs to catch everything new.
     async fn scan_repo(&self, repo: &Repo) -> Result<usize, SourceError> {
         let state_key = format!("github-repo:{}", repo.full_name);
         let prior = self.state.load(&state_key).await?;
-        let first_run = prior.is_none();
+        let since = prior.as_ref().and_then(|s| s.last_modified);
-        let max_pages = if first_run { MAX_BACKFILL_PAGES } else { 1 };
 
         let mut total = 0usize;
-        for page in 1..=max_pages {
+        let mut newest: Option<DateTime<Utc>> = since;
-            let url = format!(
+        for page in 1..=MAX_BACKFILL_PAGES {
+            let mut url = format!(
                 "https://api.github.com/repos/{}/commits?author={}&per_page={}&page={}",
                 repo.full_name, self.config.user, self.config.per_page, page
             );
+            if let Some(since_dt) = since {
+                url.push_str(&format!("&since={}", since_dt.to_rfc3339()));
+            }
             let req = self.apply_headers(self.client.get(&url));
             let resp = req
                 .send()
@@ -165,6 +279,13 @@ impl GithubRepoSource {
                 .iter()
                 .filter_map(|item| parse_commit(item, repo))
                 .collect();
+            for ev in &events {
+                newest = Some(match newest {
+                    Some(n) if ev.occurred_at > n => ev.occurred_at,
+                    Some(n) => n,
+                    None => ev.occurred_at,
+                });
+            }
             total += self.writer.upsert_events(&events).await?;
 
             if items.len() < self.config.per_page as usize {
@@ -172,7 +293,7 @@ impl GithubRepoSource {
             }
         }
 
-        self.state.touch(&state_key).await?;
+        self.state.save(&state_key, None, newest).await?;
         Ok(total)
     }
 }