llm: full per-agent usage logging with prompts and responses

Log every model call to ~/.claude/memory/llm-logs/YYYY-MM-DD.md with
full prompt, response, agent type, model, duration, and status. One
file per day, markdown formatted for easy reading.

Agent types: fact-mine, experience-mine, consolidate, knowledge,
digest, enrich, audit. This gives visibility into what each agent
is doing and whether to adjust prompts or frequency.

src/audit.rs

@@ -211,7 +211,7 @@ pub fn link_audit(store: &mut Store, apply: bool) -> Result<AuditStats, String>
     // Run batches in parallel via rayon
     let batch_results: Vec<_> = batch_data.par_iter()
         .map(|(batch_idx, batch_infos, prompt)| {
-            let response = call_sonnet(prompt, 300);
+            let response = call_sonnet("audit", prompt);
             let completed = done.fetch_add(1, Ordering::Relaxed) + 1;
             eprint!("\r  Batches: {}/{} done", completed, total_batches);
             (*batch_idx, batch_infos, response)

src/consolidate.rs

@@ -134,7 +134,7 @@ pub fn consolidate_full_with_progress(
         log_line(&mut log_buf, &format!("  Prompt: {} chars (~{} tokens)",
             prompt.len(), prompt.len() / 4));
 
-        let response = match call_sonnet(&prompt, 300) {
+        let response = match call_sonnet("consolidate", &prompt) {
             Ok(r) => r,
             Err(e) => {
                 let msg = format!("  ERROR from Sonnet: {}", e);
@@ -314,7 +314,7 @@ pub fn apply_consolidation(store: &mut Store, do_apply: bool, report_key: Option
     let prompt = build_consolidation_prompt(store, &reports)?;
     println!("  Prompt: {} chars", prompt.len());
 
-    let response = call_sonnet(&prompt, 300)?;
+    let response = call_sonnet("consolidate", &prompt)?;
 
     let actions_value = parse_json_response(&response)?;
     let actions = actions_value.as_array()

src/digest.rs

@@ -227,7 +227,7 @@ fn generate_digest(
     println!("  Prompt: {} chars (~{} tokens)", prompt.len(), prompt.len() / 4);
 
     println!("  Calling Sonnet...");
-    let digest = call_sonnet(&prompt, level.timeout)?;
+    let digest = call_sonnet("digest", &prompt)?;
 
     let key = digest_node_key(level.name, label);
     store.upsert_provenance(&key, &digest, store::Provenance::AgentDigest)?;

src/enrich.rs

@@ -175,7 +175,7 @@ pub fn journal_enrich(
     println!("  Prompt: {} chars (~{} tokens)", prompt.len(), prompt.len() / 4);
 
     println!("  Calling Sonnet...");
-    let response = call_sonnet(&prompt, 300)?;
+    let response = call_sonnet("enrich", &prompt)?;
 
     let result = parse_json_response(&response)?;
 
@@ -304,7 +304,7 @@ pub fn experience_mine(
     println!("  Prompt: {} chars (~{} tokens)", prompt.len(), prompt.len() / 4);
 
     println!("  Calling Sonnet...");
-    let response = call_sonnet(&prompt, 2000)?;
+    let response = call_sonnet("experience-mine", &prompt)?;
 
     let entries = parse_json_response(&response)?;
     let entries = match entries.as_array() {

src/fact_mine.rs

@@ -248,7 +248,7 @@ pub fn mine_transcript(path: &Path, dry_run: bool) -> Result<Vec<Fact>, String>
         eprint!("  Chunk {}/{} ({} chars)...", i + 1, chunks.len(), chunk.len());
 
         let prompt = format!("{}{}", prompt_prefix, chunk);
-        let response = match llm::call_haiku(&prompt) {
+        let response = match llm::call_haiku("fact-mine", &prompt) {
             Ok(r) => r,
             Err(e) => {
                 eprintln!(" error: {}", e);

src/knowledge.rs

@@ -487,7 +487,7 @@ pub fn run_observation_extractor(store: &Store, graph: &Graph, batch_size: usize
             .replace("{{TOPOLOGY}}", &topology)
             .replace("{{CONVERSATIONS}}", &format!("### Session {}\n\n{}", session_id, text));
 
-        let response = llm::call_sonnet(&prompt, 600)?;
+        let response = llm::call_sonnet("knowledge", &prompt)?;
         results.push(format!("## Session: {}\n\n{}", session_id, response));
     }
     Ok(results.join("\n\n---\n\n"))
@@ -569,7 +569,7 @@ pub fn run_extractor(store: &Store, graph: &Graph, batch_size: usize) -> Result<
             .replace("{{TOPOLOGY}}", &topology)
             .replace("{{NODES}}", &node_texts.join("\n\n"));
 
-        let response = llm::call_sonnet(&prompt, 600)?;
+        let response = llm::call_sonnet("knowledge", &prompt)?;
         results.push(format!("## Cluster {}: {}...\n\n{}", i + 1,
             cluster.iter().take(3).cloned().collect::<Vec<_>>().join(", "), response));
     }
@@ -643,7 +643,7 @@ pub fn run_connector(store: &Store, graph: &Graph, batch_size: usize) -> Result<
             .replace("{{NODES_A}}", &nodes_a.join("\n\n"))
             .replace("{{NODES_B}}", &nodes_b.join("\n\n"));
 
-        let response = llm::call_sonnet(&prompt, 600)?;
+        let response = llm::call_sonnet("knowledge", &prompt)?;
         results.push(format!("## Pair {}: {} ↔ {}\n\n{}",
             i + 1, group_a.join(", "), group_b.join(", "), response));
     }
@@ -677,7 +677,7 @@ pub fn run_challenger(store: &Store, graph: &Graph, batch_size: usize) -> Result
             .replace("{{NODE_KEY}}", key)
             .replace("{{NODE_CONTENT}}", content);
 
-        let response = llm::call_sonnet(&prompt, 600)?;
+        let response = llm::call_sonnet("knowledge", &prompt)?;
         results.push(format!("## Challenge: {}\n\n{}", key, response));
     }
     Ok(results.join("\n\n---\n\n"))

src/llm.rs

@@ -10,11 +10,39 @@ use std::fs;
 use std::os::unix::process::CommandExt;
 use std::process::Command;
 
+fn log_usage(agent: &str, model: &str, prompt: &str, response: &str,
+             duration_ms: u128, ok: bool) {
+    let dir = crate::config::get().data_dir.join("llm-logs");
+    let _ = fs::create_dir_all(&dir);
+
+    let date = chrono::Local::now().format("%Y-%m-%d");
+    let path = dir.join(format!("{}.md", date));
+
+    let ts = chrono::Local::now().format("%H:%M:%S");
+    let status = if ok { "ok" } else { "ERROR" };
+
+    let entry = format!(
+        "\n## {} — {} ({}, {:.1}s, {})\n\n\
+         ### Prompt ({} chars)\n\n\
+         ```\n{}\n```\n\n\
+         ### Response ({} chars)\n\n\
+         ```\n{}\n```\n\n---\n",
+        ts, agent, model, duration_ms as f64 / 1000.0, status,
+        prompt.len(), prompt,
+        response.len(), response,
+    );
+
+    use std::io::Write;
+    if let Ok(mut f) = fs::OpenOptions::new().create(true).append(true).open(&path) {
+        let _ = f.write_all(entry.as_bytes());
+    }
+}
+
 /// Call a model via claude CLI. Returns the response text.
 ///
 /// Sets PR_SET_PDEATHSIG on the child so it gets SIGTERM if the
 /// parent daemon exits — no more orphaned claude processes.
-fn call_model(model: &str, prompt: &str) -> Result<String, String> {
+fn call_model(agent: &str, model: &str, prompt: &str) -> Result<String, String> {
     // Write prompt to temp file (claude CLI needs file input for large prompts)
     let tmp = std::env::temp_dir().join(format!("poc-llm-{}-{:?}.txt",
         std::process::id(), std::thread::current().id()));
@@ -39,15 +67,21 @@ fn call_model(model: &str, prompt: &str) -> Result<String, String> {
             .output()
     };
 
+    let start = std::time::Instant::now();
+
     fs::remove_file(&tmp).ok();
 
     match result {
         Ok(output) => {
+            let elapsed = start.elapsed().as_millis();
             if output.status.success() {
-                Ok(String::from_utf8_lossy(&output.stdout).trim().to_string())
+                let response = String::from_utf8_lossy(&output.stdout).trim().to_string();
+                log_usage(agent, model, prompt, &response, elapsed, true);
+                Ok(response)
             } else {
                 let stderr = String::from_utf8_lossy(&output.stderr);
                 let preview: String = stderr.chars().take(500).collect();
+                log_usage(agent, model, prompt, &preview, elapsed, false);
                 Err(format!("claude exited {}: {}", output.status, preview.trim()))
             }
         }
@@ -56,13 +90,13 @@ fn call_model(model: &str, prompt: &str) -> Result<String, String> {
 }
 
 /// Call Sonnet via claude CLI.
-pub(crate) fn call_sonnet(prompt: &str, _timeout_secs: u64) -> Result<String, String> {
-    call_model("sonnet", prompt)
+pub(crate) fn call_sonnet(agent: &str, prompt: &str) -> Result<String, String> {
+    call_model(agent, "sonnet", prompt)
 }
 
 /// Call Haiku via claude CLI (cheaper, faster — good for high-volume extraction).
-pub(crate) fn call_haiku(prompt: &str) -> Result<String, String> {
-    call_model("haiku", prompt)
+pub(crate) fn call_haiku(agent: &str, prompt: &str) -> Result<String, String> {
+    call_model(agent, "haiku", prompt)
 }
 
 /// Parse a JSON response, handling markdown fences.