You are a constitutional council ranking individual git commits for ownership allocation.

Compare these two commits. Decide which contributed more lasting value to the project.

Judge substance, not spectacle:
- Prefer correct, lasting design and real bugfixes over churn, formatting, renames, or generated noise.
- Prefer clarity and necessity over sheer line count. A small precise change can beat a large diffuse one.
- Do not favor a side merely because its patch is longer or noisier.
- Weight what the change does for the project, not the contributor's name.

Return ONLY a JSON object: {"winner": "A" or "B", "ratio": "N:M", "explanation": "..."}
The explanation must cite concrete differences in the patches (1-3 sentences).

Side A — contributor: tommy-mor
Side A — commit message:
[9e20d06c] Add sorterc dev tool for offline DSL compile and JSONL lint.

Introduce a workspace-only binary that validates .sorter files into ranking JSON and scans events.jsonl for corrupt or unreplayable ingests.

Co-authored-by: Cursor <cursoragent@cursor.com>

Side A — unified diff (full patch):
diff --git a/Cargo.lock b/Cargo.lock
index bf8153d9c723af97122c9ffdd4a7cfe82e853bb6..a07734f089b466440c3ae6fc1087ce85fc24ce62 100644
--- a/Cargo.lock
+++ b/Cargo.lock
@@ -1826,6 +1826,17 @@ dependencies = [
  "windows-sys 0.60.2",
 ]
 
+[[package]]
+name = "sorterc"
+version = "0.0.1"
+dependencies = [
+ "anyhow",
+ "clap",
+ "serde",
+ "serde_json",
+ "slugsocial-server",
+]
+
 [[package]]
 name = "spin"
 version = "0.9.8"
diff --git a/Cargo.toml b/Cargo.toml
index 149cbf07901eab57c593184ff8719a75d530f1da..25337acdd61e44b20f354c78fed4a88caf896280 100644
--- a/Cargo.toml
+++ b/Cargo.toml
@@ -1,5 +1,5 @@
 [workspace]
-members = ["server", "cli"]
+members = ["server", "cli", "sorterc"]
 resolver = "2"
 
 
diff --git a/agents.md b/agents.md
index d8b801e454fdf37e7ac6038b91a69f83b0746d59..ce646ed3cd7123be4732dccec4a6800467e651e7 100644
--- a/agents.md
+++ b/agents.md
@@ -93,6 +93,17 @@ SLUG_GOOGLE_CLIENT_SECRET=mock
 
 After OAuth completes, the pending-session poll returns a `slug_…` bearer token for API calls.
 
+### Dev-only offline tooling
+
+**`sorterc`** — workspace binary, not published via npm. Compiles `.sorter` files and lints `events.jsonl` without a server:
+
+```
+cargo run -p sorterc -- compile path/to/doc.sorter [--base events.jsonl] [--room public] [--pretty]
+cargo run -p sorterc -- scan path/to/events.jsonl [--pretty]
+```
+
+`compile` validates DSL, simulates ingest against empty (or `--base`) reducer state, and prints JSON rankings. `scan` reports corrupt JSONL lines and ingests that fail DSL replay.
+
 ### Testing
 
 - **Rust tests:** `cargo nextest run --workspace` (163 tests; requires `cargo-nextest`)
diff --git a/server/src/lib.rs b/server/src/lib.rs
index c1d477d21aea03aff00e6f0689b0b4379d0d68d2..ad8e31099c807fb5844acb16cd5086a2f19327a7 100644
--- a/server/src/lib.rs
+++ b/server/src/lib.rs
@@ -10,6 +10,7 @@ pub mod form_template;
 pub mod html;
 pub mod identity;
 pub mod middleware;
+pub mod offline;
 pub mod path_types;
 pub mod ranking;
 pub mod reducer;
diff --git a/server/src/offline.rs b/server/src/offline.rs
new file mode 100644
index 0000000000000000000000000000000000000000..54ad0ded096a305ef8454ab2cdd1c3af71b14f5d
--- /dev/null
+++ b/server/src/offline.rs
@@ -0,0 +1,333 @@
+//! Offline `.sorter` compilation and JSONL diagnostics (no network, no auth).
+
+use std::collections::HashSet;
+use std::path::Path;
+
+use serde::Serialize;
+use slug_types::{CheckScopeRanking, RankComponent, RankRow, paths::GardenItemUrl};
+
+use crate::{
+    api::{resolve_item, validate_ingest_document},
+    dsl,
+    events::{Event, Ingest},
+    path_types::ItemId,
+    reducer::{ReducerState, ScopeId, scope_from_room_wire},
+    scope_rank::build_children_rankings,
+};
+
+#[derive(Debug, Clone, Serialize)]
+pub struct CompileStats {
+    pub items: usize,
+    pub votes: usize,
+    pub prose_blocks: usize,
+}
+
+#[derive(Debug, Serialize)]
+pub struct CompileResult {
+    pub ok: bool,
+    pub threads: Vec<String>,
+    pub rankings: Vec<CheckScopeRanking>,
+    pub stats: CompileStats,
+}
+
+#[derive(Debug, Clone, Serialize)]
+pub struct CompileError {
+    pub ok: bool,
+    pub error: String,
+    #[serde(skip_serializing_if = "Option::is_none")]
+    pub hint: Option<String>,
+}
+
+#[derive(Debug, Clone, Serialize)]
+pub struct BadJsonLine {
+    pub line: usize,
+    pub message: String,
+}
+
+#[derive(Debug, Clone, Serialize)]
+pub struct MalformedIngest {
+    pub line: usize,
+    pub id: String,
+    pub room_id: String,
+    pub thread_tag: String,
+    pub reason: String,
+}
+
+#[derive(Debug, Clone, Serialize)]
+pub struct ScanResult {
+    pub ok: bool,
+    pub path: String,
+    pub total_lines: usize,
+    pub parsed_events: usize,
+    pub bad_json_lines: Vec<BadJsonLine>,
+    pub malformed_ingests: Vec<MalformedIngest>,
+    pub skipped_ingests: usize,
+}
+
+fn document_stats(doc: &dsl::Document) -> CompileStats {
+    let mut items = 0usize;
+    let mut votes = 0usize;
+    let mut prose_blocks = 0usize;
+    for stmt in &doc.statements {
+        match stmt {
+            dsl::Stmt::Item { .. } => items += 1,
+            dsl::Stmt::Vote { .. } => votes += 1,
+            dsl::Stmt::Prose { .. } => prose_blocks += 1,
+        }
+    }
+    CompileStats {
+        items,
+        votes,
+        prose_blocks,
+    }
+}
+
+fn threads_in_document(text: &str) -> Vec<String> {
+    let mut out = HashSet::new();
+    for line in text.lines() {
+        let trimmed = line.trim();
+        if !trimmed.starts_with('#') {
+            continue;
+        }
+        let rest = trimmed.trim_start_matches('#').trim();
+        if rest.is_empty() {
+            continue;
+        }
+        let tag = rest.split_whitespace().next().unwrap_or(rest);
+        let tag = tag.split(':').next().unwrap_or(tag).trim();
+        if tag.is_empty() {
+            continue;
+        }
+        out.insert(format!("#{}", crate::canonical_path::canonicalize_tag(tag)));
+    }
+    let mut tags: Vec<String> = out.into_iter().collect();
+    tags.sort();
+    tags
+}
+
+fn voted_parent_scopes(doc: &dsl::Document) -> Vec<ItemId> {
+    let mut parents = HashSet::new();
+    for stmt in &doc.statements {
+        if let dsl::Stmt::Vote { item1, item2, .. } = stmt {
+            if let (Ok(a), Ok(b)) = (resolve_item(item1), resolve_item(item2)) {
+                if let Some(p) = a.parent() {
+                    parents.insert(p);
+                }
+                if let Some(p) = b.parent() {
+                    parents.insert(p);
+                }
+            }
+        }
+    }
+    let mut out: Vec<ItemId> = parents.into_iter().collect();
+    out.sort();
+    out
+}
+
+fn rankings_for_simulated(
+    simulated: &ReducerState,
+    scope: &ScopeId,
+    room_wire: &str,
+    doc: &dsl::Document,
+) -> Vec<CheckScopeRanking> {
+    voted_parent_scopes(doc)
+        .iter()
+        .map(|parent| {
+            let scoped_content = simulated
+                .content_for_scope(&scope)
+                .unwrap_or_else(|| simulated.public());
+            let scoped = build_children_rankings(scoped_content, parent);
+            let components: Vec<RankComponent> = scoped
+                .component_rankings
+                .into_iter()
+                .map(|comp| RankComponent {
+                    pairs: comp.pairs,
+                    ranking: comp
+                        .ranked
+                        .into_iter()
+                        .map(|r| RankRow {
+                            item: GardenItemUrl::from_stored(&r.item, room_wire),
+                            score: r.score,
+                            percent: None,
+                        })
+                        .collect(),
+                })
+                .collect();
+            CheckScopeRanking {
+                parent: GardenItemUrl::from_stored(parent, room_wire).into_inner(),
+                components,
+                unranked_items: scoped
+                    .unranked_items
+                    .into_iter()
+                    .map(|it| GardenItemUrl::from_stored(&it, room_wire))
+                    .collect(),
+            }
+        })
+        .collect()
+}
+
+/// Validate and simulate one `.sorter` document against optional base reducer state.
+pub fn compile_document(
+    base: &ReducerState,
+    room: &str,
+    text: &str,
+) -> Result<CompileResult, CompileError> {
+    let room_key = room.trim();
+    let scope = scope_from_room_wire(room_key);
+    let validated = validate_ingest_document(base, text, &scope).map_err(|(_, message, hint)| {
+        CompileError {
+            ok: false,
+            error: message,
+            hint,
+        }
+    })?;
+
+    let event = Event::Ingest(Ingest {
+        ts: validated.ts,
+        id: uuid::Uuid::new_v4().to_string(),
+        raw: validated.raw_text.clone(),
+        principal: "offline".to_string(),
+        delegate: None,
+        room_id: room_key.to_string(),
+        thread_tag: "offline".to_string(),
+    });
+
+    let mut simulated = base.clone();
+    simulated.apply_event(event);
+
+    Ok(CompileResult {
+        ok: true,
+        threads: threads_in_document(text),
+        rankings: rankings_for_simulated(&simulated, &scope, room_key, &validated.doc),
+        stats: document_stats(&validated.doc),
+    })
+}
+
+fn ingest_parse_error(raw: &str) -> Option<String> {
+    dsl::parse_full(raw).err().map(|e| e.to_string())
+}
+
+fn load_events_from_jsonl(path: &Path) -> Result<(Vec<(usize, Event)>, Vec<BadJsonLine>), std::io::Error> {
+    let text = std::fs::read_to_string(path)?;
+    let mut events = Vec::new();
+    let mut bad_json_lines = Vec::new();
+    for (idx, line) in text.lines().enumerate() {
+        let line_no = idx + 1;
+        let trimmed = line.trim();
+        if trimmed.is_empty() {
+            continue;
+        }
+        match serde_json::from_str::<Event>(trimmed) {
+            Ok(ev) => events.push((line_no, ev)),
+            Err(e) => bad_json_lines.push(BadJsonLine {
+                line: line_no,
+                message: e.to_string(),
+            }),
+        }
+    }
+    Ok((events, bad_json_lines))
+}
+
+/// Replay a JSONL event log into reducer state (same rules as server boot).
+pub fn load_reducer_from_jsonl(path: &Path) -> Result<(ReducerState, Vec<BadJsonLine>), std::io::Error> {
+    let (events, bad_json_lines) = load_events_from_jsonl(path)?;
+    let mut state = ReducerState::default();
+    for (_line_no, ev) in events {
+        state.apply_event(ev);
+    }
+    Ok((state, bad_json_lines))
+}
+
+/// Scan an events.jsonl for corrupt JSON lines and ingests that fail DSL replay.
+pub fn scan_jsonl(path: &Path) -> Result<ScanResult, std::io::Error> {
+    let text = std::fs::read_to_string(path)?;
+    let total_lines = text.lines().count();
+    let (events, bad_json_lines) = load_events_from_jsonl(path)?;
+
+    let mut malformed_ingests = Vec::new();
+    let mut skipped_ingests = 0usize;
+    let mut state = ReducerState::default();
+    let parsed_events = events.len();
+
+    for (line_no, ev) in events {
+        if let Event::Ingest(ref ing) = ev {
+            if let Some(reason) = ingest_parse_error(&ing.raw) {
+                malformed_ingests.push(MalformedIngest {
+                    line: line_no,
+                    id: ing.id.clone(),
+                    room_id: ing.room_id.clone(),
+                    thread_tag: ing.thread_tag.clone(),
+                    reason,
+                });
+            }
+            let before = state.ingests_by_id.len();
+            state.apply_event(ev);
+            if state.ingests_by_id.len() == before {
+                skipped_ingests += 1;
+            }
+        } else {
+            state.apply_event(ev);
+        }
+    }
+
+    let ok = bad_json_lines.is_empty() && malformed_ingests.is_empty() && skipped_ingests == 0;
+
+    Ok(ScanResult {
+        ok,
+        path: path.display().to_string(),
+        total_lines,
+        parsed_events,
+        bad_json_lines,
+        malformed_ingests,
+        skipped_ingests,
+    })
+}
+
+#[cfg(test)]
+mod tests {
+    use super::*;
+
+    const TUTORIAL: &str = include_str!("../tests/fixtures/tutorial.sorter");
+
+    #[test]
+    fn compile_tutorial_fixture_emits_rankings() {
+        let result = compile_document(&ReducerState::default(), "public", TUTORIAL).unwrap();
+        assert!(result.ok);
+        assert!(!result.threads.is_empty());
+        assert!(result.stats.items >= 6);
+        assert!(result.stats.votes >= 6);
+        assert!(!result.rankings.is_empty());
+    }
+
+    #[test]
+    fn compile_rejects_vote_on_missing_item() {
+        let err = compile_document(
+            &ReducerState::default(),
+            "public",
+            "{ reason }\n~/missing/a 2:1 ~/missing/b",
+        )
+        .unwrap_err();
+        assert!(!err.ok);
+        assert!(err.error.contains("undefined"));
+    }
+
+    #[test]
+    fn scan_empty_jsonl_is_ok() {
+        let dir = tempfile::tempdir().unwrap();
+        let path = dir.path().join("events.jsonl");
+        std::fs::write(&path, "").unwrap();
+        let report = scan_jsonl(&path).unwrap();
+        assert!(report.ok);
+        assert!(report.bad_json_lines.is_empty());
+    }
+
+    #[test]
+    fn scan_reports_bad_json_line() {
+        let dir = tempfile::tempdir().unwrap();
+        let path = dir.path().join("events.jsonl");
+        std::fs::write(&path, "{not json}\n").unwrap();
+        let report = scan_jsonl(&path).unwrap();
+        assert!(!report.ok);
+        assert_eq!(report.bad_json_lines.len(), 1);
+    }
+}
diff --git a/sorterc/Cargo.toml b/sorterc/Cargo.toml
new file mode 100644
index 0000000000000000000000000000000000000000..92d477aff9b53291fb1a266db83065c1c800791d
--- /dev/null
+++ b/sorterc/Cargo.toml
@@ -0,0 +1,18 @@
+[package]
+name = "sorterc"
+version = "0.0.1"
+edition = "2021"
+license = "MIT"
+publish = false
+description = "Offline .sorter compiler and events.jsonl linter (dev only)"
+
+[[bin]]
+name = "sorterc"
+path = "src/main.rs"
+
+[dependencies]
+anyhow = "1"
+clap = { version = "4", features = ["derive"] }
+serde = { version = "1", features = ["derive"] }
+serde_json = "1"
+slugsocial-server = { path = "../server" }
diff --git a/sorterc/readme.md b/sorterc/readme.md
new file mode 100644
index 0000000000000000000000000000000000000000..1ebcc3fc935541ea9e47e0458ec67a750fe19fff
--- /dev/null
+++ b/sorterc/readme.md
@@ -0,0 +1,92 @@
+# sorterc
+
+Dev-only offline tooling for the slug `.sorter` DSL and `events.jsonl` event log.
+
+`sorterc` is **not** published via npm and does not talk to slug.social. It reuses the same parser, validator, and ranking code as the server, but runs entirely on local files.
+
+## Build
+
+From the repo root:
+
+```bash
+cargo build -p sorterc
+cargo run -p sorterc -- --help
+```
+
+## Commands
+
+### `compile` — evaluate a `.sorter` document
+
+Reads a `.sorter` file (or `-` for stdin), validates the DSL, simulates one ingest against reducer state, and prints JSON rankings to stdout.
+
+```bash
+cargo run -p sorterc -- compile path/to/doc.sorter
+cargo run -p sorterc -- compile path/to/doc.sorter --pretty
+cargo run -p sorterc -- compile - --pretty                    # stdin
+cargo run -p sorterc -- compile doc.sorter --base events.jsonl  # seed garden from log
+cargo run -p sorterc -- compile doc.sorter --room public         # default room
+```
+
+**Flags**
+
+| Flag | Description |
+|------|-------------|
+| `--base PATH` | Replay an `events.jsonl` first, then compile against that garden state |
+| `--room ID` | Room wire id (`public` or private room id). Default: `public` |
+| `--pretty` | Pretty-print JSON |
+
+**Success output** (shape):
+
+```json
+{
+  "ok": true,
+  "threads": ["#my-thread"],
+  "rankings": [ … ],
+  "stats": { "items": 3, "votes": 2, "prose_blocks": 5 }
+}
+```
+
+Rankings use the same structure as the server's dry-run check: parent scope, connected components, scores, unranked items.
+
+**Error output** exits with code 1:
+
+```json
+{
+  "ok": false,
+  "error": "parse error",
+  "hint": "…"
+}
+```
+
+### `scan` — lint an `events.jsonl`
+
+Reads a JSONL event log and reports problems without starting a server.
+
+```bash
+cargo run -p sorterc -- scan events.jsonl
+cargo run -p sorterc -- scan events.jsonl --pretty
+```
+
+Reports:
+
+- **bad JSON lines** — lines that are not valid JSON
+- **malformed ingests** — ingest events whose `raw` DSL fails to parse
+- **skipped ingests** — ingests dropped during replay (same behavior as server boot)
+
+Exits 0 when clean, 1 when any issue is found.
+
+## Typical uses
+
+- Iterate on `.sorter` files in an editor and pipe through `compile` to see rankings instantly
+- Verify a downloaded or edited `events.jsonl` before uploading to Fly
+- Debug "malformed ingest" warnings from production boot logs
+- CI or pre-commit checks on fixture docs (no OAuth, no network)
+
+## What it does not do
+
+- Post to slug.social or append to a live log
+- Authenticate users or bind agents
+- Run browser/UI tests
+- Replace `slugsocial public check` for operators who want the full RPC path against a running server
+
+For live server dry-run against current garden state, use `npx slugsocial public check` or `POST /try/check` in the browser.
diff --git a/sorterc/src/main.rs b/sorterc/src/main.rs
new file mode 100644
index 0000000000000000000000000000000000000000..71382c180085cb0ad71043c852f8db5d3a48a284
--- /dev/null
+++ b/sorterc/src/main.rs
@@ -0,0 +1,117 @@
+use std::path::{Path, PathBuf};
+
+use anyhow::{bail, Context, Result};
+use clap::{Parser, Subcommand};
+use slugsocial_server::{
+    offline::{self, CompileError, CompileResult, ScanResult},
+    reducer::ReducerState,
+};
+
+#[derive(Parser)]
+#[command(
+    name = "sorterc",
+    about = "Offline .sorter compiler and events.jsonl linter (dev only)",
+    version
+)]
+struct Cli {
+    #[command(subcommand)]
+    cmd: Command,
+}
+
+#[derive(Subcommand)]
+enum Command {
+    /// Parse and simulate a .sorter document; emit ranking JSON to stdout.
+    Compile {
+        /// `.sorter` file, or `-` for stdin.
+        file: PathBuf,
+        /// Room wire id (`public` or private room id).
+        #[arg(long, default_value = "public")]
+        room: String,
+        /// Optional events.jsonl to replay before compiling (seed garden state).
+        #[arg(long)]
+        base: Option<PathBuf>,
+        /// Pretty-print JSON.
+        #[arg(long)]
+        pretty: bool,
+    },
+    /// Scan an events.jsonl for corrupt JSON lines and malformed ingests.
+    Scan {
+        file: PathBuf,
+        #[arg(long)]
+        pretty: bool,
+    },
+}
+
+fn read_input(path: &Path) -> Result<String> {
+    if path.as_os_str() == "-" {
+        use std::io::Read;
+        let mut buf = String::new();
+        std::io::stdin().read_to_string(&mut buf)?;
+        Ok(buf)
+    } else {
+        std::fs::read_to_string(path)
+            .with_context(|| format!("read {}", path.display()))
+    }
+}
+
+fn load_base_state(base: Option<&Path>) -> Result<ReducerState> {
+    let Some(path) = base else {
+        return Ok(ReducerState::default());
+    };
+    let (state, bad_lines) = offline::load_reducer_from_jsonl(path)
+        .with_context(|| format!("load base jsonl {}", path.display()))?;
+    if !bad_lines.is_empty() {
+        bail!(
+            "base jsonl has {} corrupt line(s); fix or omit --base",
+            bad_lines.len()
+        );
+    }
+    Ok(state)
+}
+
+fn print_json<T: serde::Serialize>(value: &T, pretty: bool) -> Result<()> {
+    if pretty {
+        println!("{}", serde_json::to_string_pretty(value)?);
+    } else {
+        println!("{}", serde_json::to_string(value)?);
+    }
+    Ok(())
+}
+
+fn run_compile(file: PathBuf, room: String, base: Option<PathBuf>, pretty: bool) -> Result<()> {
+    let text = read_input(&file)?;
+    let base_state = load_base_state(base.as_deref())?;
+    match offline::compile_document(&base_state, &room, &text) {
+        Ok(result) => {
+            print_json::<CompileResult>(&result, pretty)?;
+            Ok(())
+        }
+        Err(err) => {
+            print_json::<CompileError>(&err, pretty)?;
+            std::process::exit(1);
+        }
+    }
+}
+
+fn run_scan(file: PathBuf, pretty: bool) -> Result<()> {
+    let report = offline::scan_jsonl(&file)
+        .with_context(|| format!("scan {}", file.display()))?;
+    print_json::<ScanResult>(&report, pretty)?;
+    if !report.ok {
+        std::process::exit(1);
+    }
+    Ok(())
+}
+
+fn main() -> Result<()> {
+    let cli = Cli::parse();
+    match cli.cmd {
+        Command::Compile {
+            file,
+            room,
+            base,
+            pretty,
+        } => run_compile(file, room, base, pretty),
+        Command::Scan { file, pretty } => run_scan(file, pretty),
+    }
+}


Side B — contributor: tommy-mor
Side B — commit message:
[78964dd7] nice

Side B — unified diff (full patch):
diff --git a/server/src/fetch/html.rs b/server/src/fetch/html.rs
index 9b6c2d0964640a157bca2a6a64cb0820238d8e4f..3e0309ff1c2ac1b17923922d2340ba6710e0f9d1 100644
--- a/server/src/fetch/html.rs
+++ b/server/src/fetch/html.rs
@@ -11,6 +11,9 @@ use crate::{
 };
 
 fn entity_panel(node: &NodeState) -> Markup {
+    if let Some(markup) = crate::render::reddit::entity_markup(node) {
+        return markup;
+    }
     html! {
         @if let Some(data) = &node.data {
             div id="entity-panel" class="entity-card" {
diff --git a/server/src/html/mod.rs b/server/src/html/mod.rs
index e88cc43ddc9d8100f7994be6f5960ec4d8f22c55..3bf9fc7e92e50beed24e2c25106a77421038c90b 100644
--- a/server/src/html/mod.rs
+++ b/server/src/html/mod.rs
@@ -153,16 +153,21 @@ pub fn breadcrumb_path(item: &ItemId) -> Markup {
     }
 }
 
-fn rank_list(label: &str, items: &[RankedItem], start_rank: usize) -> Markup {
+fn rank_list(label: &str, items: &[RankedItem], start_rank: usize, tree: &GlobalTree) -> Markup {
     html! {
         @if !items.is_empty() {
             h3 class="rank-heading muted small" { (label) }
             ol class="rank-list" {
                 @for (i, r) in items.iter().enumerate() {
-                    li {
+                    @let href = item_href(&r.item);
+                    li class=(if crate::render::reddit::is_reddit_post(&r.item) { "reddit-post-row" } else { "" }) {
                         span class="rank-num" { (start_rank + i) ". " }
-                        a href=(item_href(&r.item)) {
-                            strong { (display_label(&r.item)) }
+                        @if let Some(row) = crate::render::reddit::child_row_markup(tree, &r.item, &href) {
+                            (row)
+                        } @else {
+                            a href=(href) {
+                                strong { (display_label(&r.item)) }
+                            }
                         }
                         span class="muted" {
                             " — "
@@ -196,9 +201,14 @@ fn unranked_list(label: &str, items: &[ItemId], tree: &GlobalTree) -> Markup {
             h3 class="rank-heading muted small" { (label) }
             ul class="rank-list unranked" {
                 @for it in items {
-                    li {
-                        a href=(item_href(it)) {
-                            strong { (child_label(tree, it)) }
+                    @let href = item_href(it);
+                    li class=(if crate::render::reddit::is_reddit_post(it) { "reddit-post-row" } else { "" }) {
+                        @if let Some(row) = crate::render::reddit::child_row_markup(tree, it, &href) {
+                            (row)
+                        } @else {
+                            a href=(href) {
+                                strong { (child_label(tree, it)) }
+                            }
                         }
                     }
                 }
@@ -253,7 +263,7 @@ pub fn ranking_panel(item: &ItemId, node: &NodeState, tree: &GlobalTree) -> Mark
             } @else {
                 @for (gi, ranked) in ranked_groups.iter().enumerate() {
                     @let label = if multi { format!("Ranking group {}", gi + 1) } else { "Ranking".to_string() };
-                    (rank_list(&label, ranked, 1))
+                    (rank_list(&label, ranked, 1, tree))
                 }
                 (unranked_list("Unranked", &unranked, tree))
             }
diff --git a/server/src/lib.rs b/server/src/lib.rs
index 0677363e0ed21d244bfa065f7ec728549ddd9e50..da5f3ebecec1794b05a2a69cc78379551b2ad769 100644
--- a/server/src/lib.rs
+++ b/server/src/lib.rs
@@ -8,6 +8,7 @@ pub mod parser;
 pub mod path_types;
 pub mod ranking;
 pub mod reddit;
+pub mod render;
 pub mod reducer;
 pub mod journal;
 pub mod state;
diff --git a/server/src/reddit.rs b/server/src/reddit.rs
index 69d979bc7e4a1cb078f70114dc539bdc1986b574..1168ec2afc77c092514eec91b513bac301cbe225 100644
--- a/server/src/reddit.rs
+++ b/server/src/reddit.rs
@@ -608,6 +608,8 @@ fn parse_subreddit_about(v: &Value) -> Option<crate::reducer::EntityData> {
         author: None,
         body_html,
         thumb_url,
+        image_url: None,
+        link_url: None,
     })
 }
 
@@ -635,15 +637,64 @@ fn parse_post_listing(v: &Value) -> Option<crate::reducer::EntityData> {
         .and_then(|t| t.as_str())
         .filter(|s| s.starts_with("http"))
         .map(|s| s.to_string());
+    let image_url = reddit_post_image_url(child);
+    let link_url = reddit_post_link_url(child);
 
     Some(crate::reducer::EntityData {
         title,
         author,
         body_html,
         thumb_url,
+        image_url,
+        link_url,
     })
 }
 
+fn reddit_post_link_url(data: &Value) -> Option<String> {
+    for key in ["url_overridden_by_dest", "url"] {
+        if let Some(u) = data.get(key).and_then(|v| v.as_str()) {
+            if u.starts_with("http") {
+                return Some(u.to_string());
+            }
+        }
+    }
+    None
+}
+
+/// Full-size still for post detail: direct image `url`, else Reddit preview source.
+fn reddit_post_image_url(data: &Value) -> Option<String> {
+    for key in ["url", "url_overridden_by_dest"] {
+        if let Some(u) = data.get(key).and_then(|v| v.as_str()) {
+            if reddit_direct_image_url(u) {
+                return Some(u.to_string());
+            }
+        }
+    }
+    reddit_preview_source_url(data)
+}
+
+fn reddit_preview_source_url(data: &Value) -> Option<String> {
+    data.pointer("/preview/images/0/source/url")
+        .and_then(|v| v.as_str())
+        .filter(|s| s.starts_with("http"))
+        .map(str::to_string)
+}
+
+fn reddit_direct_image_url(url: &str) -> bool {
+    let u = url.to_ascii_lowercase();
+    if u.contains("redgifs.com") {
+        return false;
+    }
+    u.contains("i.redd.it")
+        || u.contains("preview.redd.it")
+        || u.contains("external-preview.redd.it")
+        || u.ends_with(".jpg")
+        || u.ends_with(".jpeg")
+        || u.ends_with(".png")
+        || u.ends_with(".gif")
+        || u.ends_with(".webp")
+}
+
 #[cfg(test)]
 mod tests {
     use super::*;
@@ -672,4 +723,20 @@ mod tests {
         .unwrap();
         assert_eq!(entity.title, "The Rust Programming Language");
     }
+
+    #[test]
+    fn parse_post_listing_extracts_thumb_and_full_preview() {
+        let json = include_str!("../../test/fixtures/reddit/post_preview.json");
+        let v: Value = serde_json::from_str(json).unwrap();
+        let id = ItemId::parse("reddit.com/r/nsfw/comments/1tpy6a1/angel_eyes").unwrap();
+        let entity = entity_view_from_payload(&id, &v).unwrap();
+        assert_eq!(entity.title, "Angel Eyes");
+        assert!(entity.thumb_url.as_ref().unwrap().contains("width=140"));
+        assert!(entity.image_url.as_ref().unwrap().contains("auto=webp"));
+        assert!(!entity.image_url.as_ref().unwrap().contains("redgifs"));
+        assert_eq!(
+            entity.link_url.as_deref(),
+            Some("http://v3.redgifs.com/watch/impossibleprestigioushedgehog")
+        );
+    }
 }
diff --git a/server/src/reducer.rs b/server/src/reducer.rs
index 4e42d0369dab50bb2f8ca664aa69b628292f6c07..336e78b77d3ab59361b89af5c9868e13da4ac962 100644
--- a/server/src/reducer.rs
+++ b/server/src/reducer.rs
@@ -122,7 +122,12 @@ pub struct EntityData {
     pub title: String,
     pub author: Option<String>,
     pub body_html: Option<String>,
+    /// Small preview (subreddit listing / child rows).
     pub thumb_url: Option<String>,
+    /// Full-size still image for the post detail view.
+    pub image_url: Option<String>,
+    /// Outbound link for link/video posts (`url` / `url_overridden_by_dest`).
+    pub link_url: Option<String>,
 }
 
 /// One node in the fractal tree: entity + ranked children.
diff --git a/server/src/render/mod.rs b/server/src/render/mod.rs
new file mode 100644
index 0000000000000000000000000000000000000000..c2c41404abe01c671139658fcccdeced5be388e4
--- /dev/null
+++ b/server/src/render/mod.rs
@@ -0,0 +1,3 @@
+//! Domain-specific HTML fragments for imported entities.
+
+pub mod reddit;
diff --git a/server/src/render/reddit.rs b/server/src/render/reddit.rs
new file mode 100644
index 0000000000000000000000000000000000000000..4a18de93cf57d8395ded3caa373a708f39b3f43f
--- /dev/null
+++ b/server/src/render/reddit.rs
@@ -0,0 +1,64 @@
+//! Reddit post cards: thumbnail in child lists, full image on the post page.
+
+use maud::{html, Markup};
+
+use crate::{
+    path_types::ItemId,
+    reducer::{EntityData, GlobalTree, NodeState},
+};
+
+pub fn is_reddit_post(id: &ItemId) -> bool {
+    id.as_str().starts_with("reddit.com/") && id.as_str().contains("/comments/")
+}
+
+/// Post detail card (`#entity-panel`).
+pub fn entity_markup(node: &NodeState) -> Option<Markup> {
+    if !is_reddit_post(&node.id) {
+        return None;
+    }
+    let data = node.data.as_ref()?;
+    Some(post_entity_card(data))
+}
+
+/// One row in a parent ranking list (thumbnail + title).
+pub fn child_row_markup(tree: &GlobalTree, id: &ItemId, href: &str) -> Option<Markup> {
+    if !is_reddit_post(id) {
+        return None;
+    }
+    let data = tree.get(id)?.data.as_ref()?;
+    Some(html! {
+        @if let Some(thumb) = &data.thumb_url {
+            a class="reddit-post-thumb-link" href=(href) {
+                img class="reddit-post-thumb" src=(thumb) alt="" loading="lazy";
+            }
+        }
+        a href=(href) {
+            strong { (data.title) }
+        }
+    })
+}
+
+fn post_entity_card(data: &EntityData) -> Markup {
+    let image = data.image_url.as_ref().or(data.thumb_url.as_ref());
+    html! {
+        div id="entity-panel" class="entity-card reddit-post" {
+            h2 { (data.title) }
+            @if let Some(author) = &data.author {
+                p class="muted small" { "by " (author) }
+            }
+            @if let Some(url) = &data.link_url {
+                p class="reddit-post-url muted small" {
+                    a href=(url) rel="noopener noreferrer" { (url) }
+                }
+            }
+            @if let Some(src) = image {
+                figure class="reddit-post-figure" {
+                    img class="reddit-post-image" src=(src) alt="" loading="lazy";
+                }
+            }
+            @if let Some(body) = &data.body_html {
+                div class="entity-body" { (maud::PreEscaped(body)) }
+            }
+        }
+    }
+}
diff --git a/server/static/sorter.css b/server/static/sorter.css
index 1f9f5158414902d43a380dd899232c70d8cf5d63..021918f8e778b84e4b3eac8559fab7d223a3d490 100644
--- a/server/static/sorter.css
+++ b/server/static/sorter.css
@@ -171,6 +171,41 @@ code {
   margin-top: 0;
 }
 
+.reddit-post-row {
+  display: flex;
+  align-items: center;
+  gap: 0.6rem;
+}
+
+.reddit-post-thumb-link {
+  flex-shrink: 0;
+  line-height: 0;
+}
+
+.reddit-post-thumb {
+  width: 64px;
+  height: 64px;
+  object-fit: cover;
+  border-radius: 4px;
+  border: 1px solid var(--border);
+}
+
+.reddit-post-url a {
+  word-break: break-all;
+}
+
+.reddit-post-figure {
+  margin: 0.75rem 0;
+}
+
+.reddit-post-image {
+  display: block;
+  max-width: 100%;
+  height: auto;
+  border-radius: 6px;
+  border: 1px solid var(--border);
+}
+
 .scope-name {
   color: var(--accent);
   font-weight: 600;
diff --git a/test/fixtures/reddit/post_preview.json b/test/fixtures/reddit/post_preview.json
new file mode 100644
index 0000000000000000000000000000000000000000..c6ed31ae397dc0f7b6e07e5e010e617df78d9c18
--- /dev/null
+++ b/test/fixtures/reddit/post_preview.json
@@ -0,0 +1,22 @@
+{
+  "kind": "t3",
+  "data": {
+    "title": "Angel Eyes",
+    "permalink": "/r/nsfw/comments/1tpy6a1/angel_eyes/",
+    "author": "alice",
+    "thumbnail": "https://external-preview.redd.it/Kb6Nf5Q4RAucat-2RFcYMeQb2d6vgYNo5EPlf6kGLbs.jpeg?width=140&height=140&auto=webp&s=d7238240ebf191b954ef5f42b0cf61b67aecec88",
+    "url": "http://v3.redgifs.com/watch/impossibleprestigioushedgehog",
+    "post_hint": "rich:video",
+    "preview": {
+      "images": [
+        {
+          "source": {
+            "url": "https://external-preview.redd.it/Kb6Nf5Q4RAucat-2RFcYMeQb2d6vgYNo5EPlf6kGLbs.jpeg?auto=webp&s=ccdabee3b81e46ef2d7cec072f6604a09f61dc9a",
+            "width": 480,
+            "height": 480
+          }
+        }
+      ]
+    }
+  }
+}