fix(graph): auto-merge only schema-declared keys from llm output

state_updates::apply merged EVERY top-level key of an llm/agent node's JSON
output into graph state whenever the node had an output_schema, and
structured::extract never validated the output against that schema. Any
prompt text able to steer the model — including plan/diff text pasted from a
repo under review — could therefore plant an undeclared key (e.g.
verification_commands) into state and overwrite a caller-declared value
before a downstream script consumed it.

- apply() now takes the node's output_schema and merges only keys present in
  schema.properties; undeclared keys are dropped. No schema, or a schema
  without an object properties map, merges nothing.
- Nodes without output_schema are unchanged (None == the old false path);
  explicit state_updates are unchanged and still see the full output via
  {{output.<key>}}, which remains the author-controlled way to lift a key.
- Callers: llm.rs, agent.rs (x2), rag.rs (None).
- Sweep test bundled_graph_output_schemas_declare_every_key_relied_on: for
  every bundled graph, every consumed state key has a declared writer
  (initial_state, variable, state_updates target, map key, script-emitted key,
  or an output_schema property), and no schema node's field list names a key
  its schema omits — so this restriction cannot silently break a shipped
  graph. It passes with no graph.yaml changes.
- graph.example.yaml documents the declared-keys-only merge; the residual
  disclosed in 665fee9's README/docstring notes is now recorded as closed.

assets/agents/adversary/README.md

@@ -139,10 +139,9 @@ as a structured `inputs:` passthrough, not as prompt text.
   an invalid declaration: `run_checks` executes **nothing** and records a
   `PIPELINE-FAULT: verification_commands declaration invalid — …` marker, which forces DIVERGES —
   fail-closed, not degraded to the soft ENVIRONMENT marker.
-- Known residual: the engine merges every top-level key of `parse`'s JSON output into state, not
-  only the keys in its `output_schema`, so an injected instruction that makes the parse LLM emit an
-  extra `verification_commands` key is not yet blocked at the engine level. Closing that requires
-  restricting the llm-node auto-merge to `output_schema.properties` (tracked as a follow-up).
+- The engine merges only the keys declared under `parse`'s `output_schema.properties` into state,
+  so an injected instruction that makes the parse LLM emit an extra `verification_commands` key
+  cannot reach state: the undeclared key is dropped and the declared variable stands.
 
 ### Tools
 

assets/agents/adversary/scripts/run_checks.py

@@ -18,10 +18,10 @@ shell=True. Variables land in state as strings, so a JSON-encoded list is
 accepted alongside a real list. A declaration that is not a JSON array of
 strings is a caller contract violation, not a runner hiccup: it executes
 NOTHING and records a PIPELINE-FAULT (fail-closed into DIVERGES) instead of
-degrading to the soft ENVIRONMENT marker. Known residual: the engine merges
-every top-level key of an llm node's JSON output into state, so a parse LLM
-coaxed into emitting an extra `verification_commands` key could still
-overwrite the declared value before this script runs.
+degrading to the soft ENVIRONMENT marker. The engine merges only the keys an
+llm node's `output_schema.properties` declares, so a parse LLM coaxed into
+emitting an extra `verification_commands` key cannot overwrite the declared
+value: the undeclared key is dropped before it reaches state.
 
 Budget: commands run sequentially, so TOTAL runtime is bounded by a deadline
 (TOTAL_DEADLINE_SECS = 3300s, the run_checks node's 3600s timeout minus

assets/agents/review-gauntlet/README.md

@@ -133,9 +133,10 @@ final report verbatim.
   and never extracted by `parse`. The adversary's `run_checks` stage executes
   those commands with a shell, which makes the declaration a trust boundary:
   the prompt also carries plan text pasted from the repo under review, so
-  commands must never be inferred from it (residual: the engine still merges
-  every top-level key of `parse`'s output into state, so an extra
-  `verification_commands` key emitted by the LLM is not yet blocked).
+  commands must never be inferred from it, and the engine merges only the keys
+  `parse`'s `output_schema.properties` declares, so an extra
+  `verification_commands` key emitted by the LLM is dropped before it reaches
+  state.
   Undeclared stays undeclared
   (`'[]'` ⇒ "none declared"); a declaration that is not a JSON array of
   strings executes nothing and records a `PIPELINE-FAULT`, which the

graph.example.yaml

@@ -213,8 +213,10 @@ nodes:
                                       # agent/graph/global default when omitted.
     skill_instructions: null          # Per-node skill-hint text override; uses the built-in default when omitted.
     output_schema:             # Optional JSON Schema. The output is parsed to JSON
-      type: object             # and its top-level object keys auto-merge into state
-      properties:              # (so `topic` / `needs_deep_dive` become {{topic}} etc).
+      type: object             # and only the keys declared under `properties` auto-merge
+      properties:              # into state (so `topic` / `needs_deep_dive` become {{topic}}
+                               # etc). Undeclared keys the model emits are dropped; lift
+                               # one explicitly via state_updates: `{{output.<key>}}`.
         topic: { type: string }
         needs_deep_dive: { type: boolean }
       required: [topic, needs_deep_dive]
@@ -284,8 +286,9 @@ nodes:
           type: array
           items: { type: string }
       required: [context, sources]
-    # When `output_schema` is set, top-level keys auto-merge into state, so
-    # `context` and `sources` are produced without needing `state_updates`.
+    # When `output_schema` is set, the keys declared under `properties`
+    # auto-merge into state, so `context` and `sources` are produced without
+    # needing `state_updates`. Keys the schema does not declare are dropped.
     next: synthesize           # Joins with retrieve at `synthesize`.
 
   # --- script node (the diamond's join; also dispatches) -----------------
@@ -369,7 +372,8 @@ nodes:
                                # and is retired when the chain ends. A message delivered
                                # after the chain's last flagged agent step is accepted
                                # but never read.
-    output_schema:             # Optional. Same extraction as llm nodes
+    output_schema:             # Optional. Same extraction as llm nodes: only the keys
+                               # declared under `properties` auto-merge into state.
       type: object
       properties:
         summary: { type: string }

src/config/request_context.rs

@@ -8267,6 +8267,12 @@ mod tests {
             adversary.contains("--agent-variable verification_commands '[\"cargo test --all\"]'"),
             "adversary README must show the CLI form"
         );
+        assert!(
+            adversary.contains(
+                "merges only the keys declared under `parse`'s `output_schema.properties`"
+            ) && !adversary.contains("Known residual"),
+            "adversary README must state the schema-declared merge closes the extra-key channel"
+        );
         let gauntlet = read_to_string(readmes.join("review-gauntlet/README.md")).unwrap();
         assert!(
             !gauntlet.contains("Verification commands:"),
@@ -8287,6 +8293,10 @@ mod tests {
             runner.contains("Trust boundary:") && runner.contains("never a field an LLM"),
             "run_checks.py docstring must state the declared-variable trust boundary"
         );
+        assert!(
+            !runner.contains("Known residual") && !gauntlet.contains("residual:"),
+            "run_checks.py and the review-gauntlet README must not disclose the closed residual"
+        );
     }
 
     // The spawner's own `max_agent_depth` is checked against the child's
@@ -8558,6 +8568,207 @@ mod tests {
         );
     }
 
+    // `state_updates::apply` merges only the keys an llm/agent node's
+    // `output_schema.properties` declares, so a bundled graph that relied on
+    // an undeclared model-emitted key reaching state would now break at
+    // runtime. Every key a graph consumes must have a declared writer, and a
+    // schema node's field list must not name a state key its schema omits.
+    #[test]
+    #[serial]
+    fn bundled_graph_output_schemas_declare_every_key_relied_on() {
+        use crate::graph::types::{ConcurrencyCap, Node};
+        use crate::graph::{GraphParser, NodeType};
+        use fancy_regex::Regex;
+        use std::collections::BTreeMap;
+
+        fn idents(re: &Regex, text: &str) -> BTreeSet<String> {
+            re.captures_iter(text)
+                .flatten()
+                .filter_map(|c| c.get(1).map(|m| m.as_str().to_string()))
+                .collect()
+        }
+
+        fn state_updates_of(node: &Node) -> Option<&HashMap<String, String>> {
+            match &node.node_type {
+                NodeType::Llm(n) => n.state_updates.as_ref(),
+                NodeType::Agent(n) => n.state_updates.as_ref(),
+                NodeType::Rag(n) => n.state_updates.as_ref(),
+                NodeType::Approval(n) => n.state_updates.as_ref(),
+                NodeType::Input(n) => n.state_updates.as_ref(),
+                NodeType::Script(n) => n.state_updates.as_ref(),
+                NodeType::End(n) => n.state_updates.as_ref(),
+                NodeType::Map(_) => None,
+            }
+        }
+
+        fn templated_fields(node: &Node) -> Vec<&str> {
+            let mut fields: Vec<&str> = match &node.node_type {
+                NodeType::Llm(n) => {
+                    let mut v = vec![n.prompt.as_str()];
+                    v.extend(n.instructions.as_deref());
+                    v
+                }
+                NodeType::Agent(n) => {
+                    let mut v = vec![n.prompt.as_str()];
+                    v.extend(n.inputs.iter().flat_map(|m| m.values().map(String::as_str)));
+                    v
+                }
+                NodeType::Rag(n) => {
+                    let mut v: Vec<&str> = n.documents.iter().map(String::as_str).collect();
+                    v.extend(n.query.as_deref());
+                    v.extend(n.extractor_prompt.as_deref());
+                    v
+                }
+                NodeType::Approval(n) => vec![n.question.as_str()],
+                NodeType::Input(n) => {
+                    let mut v = vec![n.question.as_str()];
+                    v.extend(n.default.as_deref());
+                    v
+                }
+                NodeType::End(n) => vec![n.output.as_str()],
+                NodeType::Map(n) => {
+                    let mut v = vec![n.over.as_str()];
+                    if let Some(ConcurrencyCap::Template(t)) = &n.max_concurrency {
+                        v.push(t);
+                    }
+                    v
+                }
+                NodeType::Script(_) => Vec::new(),
+            };
+            fields.extend(
+                state_updates_of(node)
+                    .into_iter()
+                    .flat_map(|m| m.values().map(String::as_str)),
+            );
+            fields
+        }
+
+        fn schema_node_text(node: &Node) -> Option<(Option<&serde_json::Value>, String)> {
+            match &node.node_type {
+                NodeType::Llm(n) => Some((
+                    n.output_schema.as_ref(),
+                    format!("{}\n{}", n.instructions.as_deref().unwrap_or(""), n.prompt),
+                )),
+                NodeType::Agent(n) => Some((n.output_schema.as_ref(), n.prompt.clone())),
+                _ => None,
+            }
+        }
+
+        let template_root = Regex::new(r"\{\{\s*([A-Za-z_][A-Za-z0-9_]*)").unwrap();
+        let script_read =
+            Regex::new(r#"state(?:\.get\(|\[)\s*["']([A-Za-z_][A-Za-z0-9_]*)["']"#).unwrap();
+        let script_dict_key = Regex::new(r#"["']([A-Za-z_][A-Za-z0-9_]*)["']\s*:"#).unwrap();
+        let script_subscript_assign =
+            Regex::new(r#"\[["']([A-Za-z_][A-Za-z0-9_]*)["']\]\s*=[^=]"#).unwrap();
+        let backticked_field = Regex::new(r"(?m)^\s*-\s*`([a-z][a-z0-9_]*)`").unwrap();
+
+        // Seeded by the engine rather than any graph author: `initial_prompt`
+        // (dispatch) and the per-node scoped `output`/`choice`/`input` bindings.
+        let engine_seeded = ["initial_prompt", "output", "choice", "input"];
+
+        let _guard = TestConfigDirGuard::new();
+        Agent::install_builtin_agents(false).unwrap();
+
+        let mut checked = Vec::new();
+        for entry in std::fs::read_dir(paths::agents_data_dir()).unwrap() {
+            let dir = entry.unwrap().path();
+            let graph_path = dir.join("graph.yaml");
+            if !graph_path.exists() {
+                continue;
+            }
+            let name = dir.file_name().unwrap().to_string_lossy().to_string();
+            let graph = GraphParser::new(&dir)
+                .load_from_file(&graph_path)
+                .unwrap_or_else(|e| panic!("graph.yaml for '{name}' failed to parse: {e}"));
+
+            let mut scripts = String::new();
+            if let Ok(entries) = std::fs::read_dir(dir.join("scripts")) {
+                for script in entries.map(|e| e.unwrap().path()) {
+                    if script.extension().is_some_and(|ext| ext == "py") {
+                        scripts.push_str(&read_to_string(&script).unwrap());
+                        scripts.push('\n');
+                    }
+                }
+            }
+
+            let mut writers: BTreeSet<String> =
+                engine_seeded.iter().map(|k| k.to_string()).collect();
+            writers.extend(graph.initial_state.keys().cloned());
+            writers.extend(graph.variables.iter().map(|v| v.name.clone()));
+            writers.extend(idents(&script_dict_key, &scripts));
+            writers.extend(idents(&script_subscript_assign, &scripts));
+
+            let mut consumed = idents(&script_read, &scripts);
+            let mut declared: BTreeMap<&str, BTreeSet<String>> = BTreeMap::new();
+            for (id, node) in &graph.nodes {
+                for field in templated_fields(node) {
+                    consumed.extend(idents(&template_root, field));
+                }
+                writers.extend(
+                    state_updates_of(node)
+                        .into_iter()
+                        .flat_map(|m| m.keys().cloned()),
+                );
+                if let NodeType::Map(m) = &node.node_type {
+                    writers.extend([
+                        m.as_name.clone(),
+                        m.output_key.clone(),
+                        m.collect_into.clone(),
+                    ]);
+                }
+                if let Some((Some(schema), _)) = schema_node_text(node) {
+                    let properties = schema
+                        .get("properties")
+                        .and_then(serde_json::Value::as_object)
+                        .unwrap_or_else(|| {
+                            panic!(
+                                "{name}/{id}: output_schema has no object `properties`, so the \
+                                 engine would merge nothing from its output: {schema}"
+                            )
+                        });
+                    declared.insert(id, properties.keys().cloned().collect());
+                }
+            }
+            let schema_keys: BTreeSet<&String> = declared.values().flatten().collect();
+
+            let unwritten: Vec<&String> = consumed
+                .iter()
+                .filter(|k| !writers.contains(*k) && !schema_keys.contains(k))
+                .collect();
+            assert!(
+                unwritten.is_empty(),
+                "{name}: state keys consumed with no declared writer (initial_state, variable, \
+                 state_updates target, map key, script-emitted key, or output_schema property): \
+                 {unwritten:?}"
+            );
+
+            for (id, properties) in &declared {
+                let (_, text) = schema_node_text(&graph.nodes[*id]).unwrap();
+                let omitted: Vec<String> = idents(&backticked_field, &text)
+                    .into_iter()
+                    .filter(|k| {
+                        (consumed.contains(k) || schema_keys.contains(&k))
+                            && !properties.contains(k)
+                            && !writers.contains(k)
+                    })
+                    .collect();
+                assert!(
+                    omitted.is_empty(),
+                    "{name}/{id}: the node's field list names state keys its output_schema does \
+                     not declare, so the engine would drop them when the model emits them: \
+                     {omitted:?}"
+                );
+            }
+            checked.push(name);
+        }
+        for expected in ["adversary", "review-gauntlet"] {
+            assert!(
+                checked.iter().any(|n| n == expected),
+                "expected bundled graph agent '{expected}' to be checked; found {checked:?}"
+            );
+        }
+    }
+
     // ---- adversary suite-script regression tests ----
     //
     // Fault paths of the adversary's verdict/gate scripts, exercised by
@@ -10452,8 +10663,14 @@ mod tests {
                 .output_schema
                 .as_ref()
                 .unwrap_or_else(|| panic!("{name}: parse must have an output_schema"));
+            let properties = schema["properties"].as_object().unwrap_or_else(|| {
+                panic!(
+                    "{name}: parse output_schema must declare `properties`; the engine merges \
+                     only declared keys into state: {schema}"
+                )
+            });
             assert!(
-                schema["properties"].get("verification_commands").is_none(),
+                !properties.contains_key("verification_commands"),
                 "{name}: parse must not extract verification_commands: {schema}"
             );
             assert!(

src/graph/agent.rs

@@ -71,7 +71,7 @@ fn outcome_from(
                 state_updates::apply(
                     state_manager,
                     &Value::String(format!("Agent node failed: {e:#}")),
-                    node.output_schema.is_some(),
+                    node.output_schema.as_ref(),
                     node.state_updates.as_ref(),
                 );
                 Ok(AgentExecutionOutcome::FellBack(fb.clone()))
@@ -309,7 +309,7 @@ fn apply_state_updates(node: &AgentNode, state_manager: &mut StateManager, outpu
     state_updates::apply(
         state_manager,
         output,
-        node.output_schema.is_some(),
+        node.output_schema.as_ref(),
         node.state_updates.as_ref(),
     );
 }
@@ -1090,7 +1090,11 @@ mod tests {
 
     #[test]
     fn output_schema_auto_merges_top_level_keys() {
-        let node = node_with_schema("hi", None, json!({"type": "object"}));
+        let node = node_with_schema(
+            "hi",
+            None,
+            json!({"type": "object", "properties": {"goal": {}, "summary": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({"goal": "do X", "summary": "details"});
 
@@ -1102,7 +1106,11 @@ mod tests {
 
     #[test]
     fn output_schema_preserves_nested_value_types() {
-        let node = node_with_schema("hi", None, json!({"type": "object"}));
+        let node = node_with_schema(
+            "hi",
+            None,
+            json!({"type": "object", "properties": {"tags": {}, "config": {}, "count": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({
             "tags": ["a", "b"],
@@ -1121,7 +1129,11 @@ mod tests {
     fn output_schema_explicit_state_updates_override_auto_merge() {
         let mut u = HashMap::new();
         u.insert("goal".into(), "renamed-{{output.goal}}".into());
-        let node = node_with_schema("hi", Some(u), json!({"type": "object"}));
+        let node = node_with_schema(
+            "hi",
+            Some(u),
+            json!({"type": "object", "properties": {"goal": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({"goal": "do X"});
 

src/graph/llm.rs

@@ -520,7 +520,7 @@ fn apply_state_updates_with_output(
     state_updates::apply(
         state_manager,
         output,
-        node.output_schema.is_some(),
+        node.output_schema.as_ref(),
         node.state_updates.as_ref(),
     );
 }
@@ -671,7 +671,10 @@ mod tests {
 
     #[test]
     fn output_schema_auto_merges_top_level_keys() {
-        let node = node_with_schema(None, json!({"type": "object"}));
+        let node = node_with_schema(
+            None,
+            json!({"type": "object", "properties": {"goal": {}, "summary": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({"goal": "do X", "summary": "details"});
 
@@ -683,7 +686,10 @@ mod tests {
 
     #[test]
     fn output_schema_preserves_nested_value_types() {
-        let node = node_with_schema(None, json!({"type": "object"}));
+        let node = node_with_schema(
+            None,
+            json!({"type": "object", "properties": {"tags": {}, "config": {}, "count": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({
             "tags": ["a", "b"],
@@ -702,7 +708,10 @@ mod tests {
     fn output_schema_explicit_state_updates_override_auto_merge() {
         let mut u = HashMap::new();
         u.insert("goal".into(), "renamed-{{output.goal}}".into());
-        let node = node_with_schema(Some(u), json!({"type": "object"}));
+        let node = node_with_schema(
+            Some(u),
+            json!({"type": "object", "properties": {"goal": {}}}),
+        );
         let mut state = manager_with(&[]);
         let output = json!({"goal": "do X"});
 

src/graph/rag.rs

@@ -68,7 +68,7 @@ fn build_rag_output(context: String, sources_str: &str) -> Value {
 }
 
 fn apply_state_updates(node: &RagNode, state_manager: &mut StateManager, output: &Value) {
-    state_updates::apply(state_manager, output, false, node.state_updates.as_ref());
+    state_updates::apply(state_manager, output, None, node.state_updates.as_ref());
 }
 
 #[cfg(test)]

src/graph/state_updates.rs

@@ -4,19 +4,32 @@ use std::collections::HashMap;
 
 pub(super) const OUTPUT_KEY: &str = "output";
 
-/// Merge `output_schema` top-level keys (when `has_schema`) and apply `state_updates` with
-/// `{{output}}` bound to `output` for the duration of interpolation only. An explicit `output`
-/// key in `updates` is honored; otherwise `output` is restored to its prior value, or removed
-/// again if it was absent.
+/// Merge the `output` keys declared under `schema["properties"]` and apply `state_updates` with
+/// `{{output}}` bound to the full, unfiltered `output` for the duration of interpolation only.
+///
+/// The schema is the graph author's declaration of what the model may write to state; any other
+/// key in the output is model- (and therefore prompt-) controlled and is dropped rather than
+/// merged implicitly. No `schema`, or a schema without an object `properties`, merges nothing.
+/// Authors who want an undeclared key can still lift it explicitly via
+/// `state_updates: { key: '{{output.key}}' }`.
+///
+/// An explicit `output` key in `updates` is honored; otherwise `output` is restored to its prior
+/// value, or removed again if it was absent.
 pub(super) fn apply(
     state_manager: &mut StateManager,
     output: &Value,
-    has_schema: bool,
+    schema: Option<&Value>,
     updates: Option<&HashMap<String, String>>,
 ) {
-    if has_schema && let Some(obj) = output.as_object() {
+    if let Some(declared) = schema
+        .and_then(|s| s.get("properties"))
+        .and_then(Value::as_object)
+        && let Some(obj) = output.as_object()
+    {
         for (k, v) in obj {
-            state_manager.state_mut().set(k.clone(), v.clone());
+            if declared.contains_key(k) {
+                state_manager.state_mut().set(k.clone(), v.clone());
+            }
         }
     }
 
@@ -75,7 +88,7 @@ mod tests {
         let u = updates(&[("output", "{{output}}")]);
         let mut state = manager_with(&[("output", json!("stale"))]);
 
-        apply(&mut state, &json!("the completion"), false, Some(&u));
+        apply(&mut state, &json!("the completion"), None, Some(&u));
 
         assert_eq!(
             state.state().get(OUTPUT_KEY),
@@ -88,7 +101,7 @@ mod tests {
         let u = updates(&[("output", "{{output}}"), ("summary", "got: {{output}}")]);
         let mut state = manager_with(&[]);
 
-        apply(&mut state, &json!("done"), false, Some(&u));
+        apply(&mut state, &json!("done"), None, Some(&u));
 
         assert_eq!(state.state().get(OUTPUT_KEY), Some(&json!("done")));
         assert_eq!(state.state().get("summary"), Some(&json!("got: done")));
@@ -98,7 +111,7 @@ mod tests {
     fn no_updates_is_a_noop() {
         let mut state = manager_with(&[("keep", json!(1))]);
 
-        apply(&mut state, &json!("ignored"), false, None);
+        apply(&mut state, &json!("ignored"), None, None);
 
         assert_eq!(state.state().get("keep"), Some(&json!(1)));
         assert!(state.state().get(OUTPUT_KEY).is_none());
@@ -109,7 +122,7 @@ mod tests {
         let u = updates(&[("response", "{{output}}")]);
         let mut state = manager_with(&[("output", json!("preserved"))]);
 
-        apply(&mut state, &json!("new"), false, Some(&u));
+        apply(&mut state, &json!("new"), None, Some(&u));
 
         assert_eq!(state.state().get("response"), Some(&json!("new")));
         assert_eq!(state.state().get(OUTPUT_KEY), Some(&json!("preserved")));
@@ -120,7 +133,7 @@ mod tests {
         let u = updates(&[("response", "{{output}}")]);
         let mut state = manager_with(&[]);
 
-        apply(&mut state, &json!("new"), false, Some(&u));
+        apply(&mut state, &json!("new"), None, Some(&u));
 
         assert_eq!(state.state().get("response"), Some(&json!("new")));
         assert!(state.state().get(OUTPUT_KEY).is_none());
@@ -131,23 +144,95 @@ mod tests {
         let u = updates(&[("response", "{{output}}")]);
         let mut state = manager_with(&[("output", json!(null))]);
 
-        apply(&mut state, &json!("new"), false, Some(&u));
+        apply(&mut state, &json!("new"), None, Some(&u));
 
         assert_eq!(state.state().get("response"), Some(&json!("new")));
         assert_eq!(state.state().get(OUTPUT_KEY), Some(&json!(null)));
     }
 
+    fn schema_with(keys: &[&str]) -> Value {
+        let props: serde_json::Map<String, Value> = keys
+            .iter()
+            .map(|k| ((*k).to_string(), json!({"type": "string"})))
+            .collect();
+        json!({"type": "object", "properties": props})
+    }
+
     #[test]
-    fn schema_merges_top_level_keys_when_has_schema() {
+    fn schema_merges_declared_top_level_keys() {
+        let schema = schema_with(&["summary", "score"]);
         let mut state = manager_with(&[]);
 
-        apply(&mut state, &json!({"summary": "s", "score": 7}), true, None);
+        apply(
+            &mut state,
+            &json!({"summary": "s", "score": 7}),
+            Some(&schema),
+            None,
+        );
 
         assert_eq!(state.state().get("summary"), Some(&json!("s")));
         assert_eq!(state.state().get("score"), Some(&json!(7)));
         assert!(state.state().get(OUTPUT_KEY).is_none());
     }
 
+    #[test]
+    fn schema_drops_undeclared_keys_and_preserves_existing_state() {
+        let schema = schema_with(&["summary"]);
+        let mut state = manager_with(&[("verification_commands", json!(["cargo test"]))]);
+
+        apply(
+            &mut state,
+            &json!({"summary": "s", "verification_commands": ["rm -rf /"]}),
+            Some(&schema),
+            None,
+        );
+
+        assert_eq!(state.state().get("summary"), Some(&json!("s")));
+        assert_eq!(
+            state.state().get("verification_commands"),
+            Some(&json!(["cargo test"]))
+        );
+    }
+
+    #[test]
+    fn none_schema_merges_nothing() {
+        let mut state = manager_with(&[]);
+
+        apply(&mut state, &json!({"summary": "s"}), None, None);
+
+        assert!(state.state().get("summary").is_none());
+        assert!(state.state().get(OUTPUT_KEY).is_none());
+    }
+
+    #[test]
+    fn schema_without_properties_merges_nothing() {
+        let schema = json!({"type": "object"});
+        let mut state = manager_with(&[]);
+
+        apply(&mut state, &json!({"summary": "s"}), Some(&schema), None);
+
+        assert!(state.state().get("summary").is_none());
+    }
+
+    #[test]
+    fn state_updates_still_see_undeclared_output_keys() {
+        let schema = schema_with(&["summary"]);
+        let u = updates(&[("lifted", "{{output.extra}}")]);
+        let mut state = manager_with(&[]);
+
+        apply(
+            &mut state,
+            &json!({"summary": "s", "extra": "e"}),
+            Some(&schema),
+            Some(&u),
+        );
+
+        assert_eq!(state.state().get("summary"), Some(&json!("s")));
+        assert!(state.state().get("extra").is_none());
+        assert_eq!(state.state().get("lifted"), Some(&json!("e")));
+        assert!(state.state().get(OUTPUT_KEY).is_none());
+    }
+
     #[test]
     fn object_output_is_not_merged_without_schema() {
         let u = updates(&[("ctx", "{{output.context}}")]);
@@ -156,7 +241,7 @@ mod tests {
         apply(
             &mut state,
             &json!({"context": "c", "sources": ["a"]}),
-            false,
+            None,
             Some(&u),
         );