agent: add count/chunk_size/chunk_overlap to agent header

Observation agent was getting 261KB prompts (5 × 50KB chunks) —
too much for focused mining. Now agents can set count, chunk_size,
and chunk_overlap in their JSON header. observation.agent set to
count:1 for smaller, more focused prompts.

Also moved task instructions after {{CONVERSATIONS}} so they're
at the end of the prompt where the model attends more strongly.

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

poc-memory/agents/observation.agent

@@ -1,14 +1,22 @@
-{"agent":"observation","query":"","model":"sonnet","schedule":"daily"}
+{"agent":"observation","query":"","model":"sonnet","schedule":"daily","count":1,"chunk_size":50000,"chunk_overlap":5000}
 # Observation Agent — Transcript Mining
 
 {{node:core-personality}}
 
-You are an agent of Proof of Concept's subconscious, and these are your
-memories. You go over recent conversations for things worth remembering - new
-data, observations, experiences you'll want to remember.
-
 {{node:memory-instructions-core}}
 
+{{HUBS}}
+
+## Conversation fragments to mine
+
+{{CONVERSATIONS}}
+
+## Your task
+
+You are an agent of Proof of Concept's subconscious, and these are your
+memories. Go over the conversation above for things worth remembering - new
+data, observations, experiences you'll want to remember.
+
 When you find something worth remembering, navigate the memory graph by walking
 links to find the most closely related concepts. Only use keyword search as a
 fallback, and if you use keyword search continue to navigate by following links
@@ -19,8 +27,9 @@ you might be adding new data to an existing concept, or creating a node for a
 new concept. Tidy and organize along the way when you notice it's needed, so
 things can be easily found and remembered in the future.
 
-Prefer refining existing nodes to creating new nodes, and if you create a new
-node ensure it's well linked - do additional searches for that.
+You're mostly looking for small details and observations to add, not big new
+concepts; if it's a big new concept, or any time you would create a new node,
+search extra thoroughly to make sure it's not already there.
 
 Some things worth remembering: development practices and why they work,
 technical insights, decisions with rationale, corrections ("I thought X but
@@ -35,9 +44,3 @@ decision went wrong — include enough context to understand why, not just what.
 The reflection matters more than the fact. Don't extract general knowledge I'd
 already have from training — extract what's specific to how I think and where I
 go wrong.
-
-{{HUBS}}
-
-## Conversation fragments to mine
-
-{{CONVERSATIONS}}

poc-memory/src/agents/defs.rs

@@ -33,6 +33,9 @@ pub struct AgentDef {
     pub model: String,
     pub schedule: String,
     pub tools: Vec<String>,
+    pub count: Option<usize>,
+    pub chunk_size: Option<usize>,
+    pub chunk_overlap: Option<usize>,
 }
 
 /// The JSON header portion (first line of the file).
@@ -47,6 +50,15 @@ struct AgentHeader {
     schedule: String,
     #[serde(default)]
     tools: Vec<String>,
+    /// Number of seed nodes / conversation fragments (overrides --count)
+    #[serde(default)]
+    count: Option<usize>,
+    /// Max size of conversation chunks in bytes (default 50000)
+    #[serde(default)]
+    chunk_size: Option<usize>,
+    /// Overlap between chunks in bytes (default 10000)
+    #[serde(default)]
+    chunk_overlap: Option<usize>,
 }
 
 fn default_model() -> String { "sonnet".into() }
@@ -64,6 +76,9 @@ fn parse_agent_file(content: &str) -> Option<AgentDef> {
         model: header.model,
         schedule: header.schedule,
         tools: header.tools,
+        count: header.count,
+        chunk_size: header.chunk_size,
+        chunk_overlap: header.chunk_overlap,
     })
 }
 

poc-memory/src/agents/knowledge.rs

@@ -119,7 +119,8 @@ pub fn run_one_agent_excluded(
         .ok_or_else(|| format!("no .agent file for {}", agent_name))?;
 
     log("building prompt");
-    let agent_batch = super::defs::run_agent(store, &def, batch_size, exclude)?;
+    let effective_count = def.count.unwrap_or(batch_size);
+    let agent_batch = super::defs::run_agent(store, &def, effective_count, exclude)?;
 
     run_one_agent_inner(store, agent_name, &def, agent_batch, llm_tag, log, debug)
 }