ei-tui 0.5.4 → 0.6.1

package/src/core/utils/levenshtein.ts

@@ -0,0 +1,18 @@
+export function levenshtein(a: string, b: string): number {
+  const m = a.length, n = b.length;
+  const dp: number[][] = Array.from({ length: m + 1 }, (_, i) =>
+    Array.from({ length: n + 1 }, (_, j) => (i === 0 ? j : j === 0 ? i : 0))
+  );
+  for (let i = 1; i <= m; i++) {
+    for (let j = 1; j <= n; j++) {
+      dp[i][j] = a[i-1] === b[j-1]
+        ? dp[i-1][j-1]
+        : 1 + Math.min(dp[i-1][j], dp[i][j-1], dp[i-1][j-1]);
+    }
+  }
+  return dp[m][n];
+}
+
+export function normalizeForMatch(s: string): string {
+  return s.toLowerCase().trim().replace(/[^\w\s]/g, '');
+}

package/src/integrations/claude-code/importer.ts

@@ -143,6 +143,7 @@ function ensureSessionTopic(
     persona_groups: CLAUDE_CODE_TOPIC_GROUPS,
     learned_by: stateManager.persona_getByName(CLAUDE_CODE_PERSONA_NAME)?.id ?? undefined,
     last_updated: new Date().toISOString(),
+    learned_on: new Date().toISOString(),
   };
 
   stateManager.human_topic_upsert(newTopic);

package/src/integrations/cursor/importer.ts

@@ -127,6 +127,7 @@ function ensureSessionTopic(
     persona_groups: CURSOR_TOPIC_GROUPS,
     learned_by: stateManager.persona_getByName(CURSOR_PERSONA_NAME)?.id ?? undefined,
     last_updated: new Date().toISOString(),
+    learned_on: new Date().toISOString(),
   };
 
   stateManager.human_topic_upsert(newTopic);

package/src/prompts/ceremony/index.ts

@@ -4,6 +4,7 @@ export { buildDescriptionCheckPrompt } from "./description-check.js";
 export { buildRewriteScanPrompt, buildRewritePrompt } from "./rewrite.js";
 export { buildDedupPrompt } from "./dedup.js";
 export { buildUserDedupPrompt } from "./user-dedup.js";
+export { buildPersonMigrationPrompt, type PersonMigrationPromptData } from "./person-migration.js";
 export type {
   PersonaExpirePromptData,
   PersonaExpireResult,

package/src/prompts/ceremony/person-migration.ts

@@ -0,0 +1,77 @@
+import type { Person } from "../../core/types/data-items.js";
+import type { PromptOutput } from "../persona/types.js";
+
+export interface PersonMigrationPromptData {
+  person: Pick<Person, "name" | "description" | "relationship">;
+}
+
+export function buildPersonMigrationPrompt(data: PersonMigrationPromptData): PromptOutput {
+  const { person } = data;
+
+  const system = `You are extracting identifiers for a Person record in a personal knowledge system.
+
+A Person record has a \`name\` and \`description\` that describe who they are. Your job is to extract ALL identifiers for this person — every name, handle, alias, or platform ID that refers to them.
+
+## CRITICAL RULES
+
+1. The \`name\` field value MUST appear in identifiers — never lose it.
+2. If \`name\` contains a space (e.g. "Jeremy Scherer"), create BOTH a \`Full Name\` identifier AND a \`First Name\` identifier for the given name.
+3. Mark exactly one identifier as \`is_primary: true\` — the most natural display name.
+4. Check \`read_memory\` for additional context before finalizing.
+5. Return ONLY a JSON object with an \`identifiers\` array. No other text.
+6. "none" is never valid — every person has at least one identifier (their name).
+
+## DESCRIPTION PRE-PROCESSING
+
+If the description begins with a JSON block (e.g. \`{"identifiers": [...]}\` or \`[{...}]\`), those are pre-seeded identifiers from a prior migration step:
+- Parse them out and include them in your output
+- Normalize their \`type\` values to Title Case (e.g. \`nickname\` → \`Nickname\`, \`full_name\` → \`Full Name\`)
+- Treat the remainder of the description (after the JSON block) as the actual description text for \`read_memory\` context
+
+## IDENTIFIER TYPES
+
+Use Title Case for all types. Built-in types:
+
+| Type | Meaning |
+|------|---------|
+| Full Name | Legal or full birth name |
+| First Name | Given/first name only |
+| Nickname | Informal name, diminutive, pet name |
+| Email | Email address |
+| GitHub | GitHub username |
+| Discord | Discord username |
+| Roblox | Roblox username |
+| Reddit | Reddit username |
+| Twitter | Twitter/X handle |
+| FF14 | Final Fantasy XIV character name |
+| Relationship | How the user addresses this person: Dad, Pop, Sis, etc. |
+| Ei Persona | Ei persona UUID (only if explicitly in the data) |
+
+Any string is valid as a type — users define their own (e.g. \`Slack-ASU\`, \`Slack-RnP\`, \`sehimu_thinara\`). Use the exact casing the user would recognize.
+
+## RESPONSE FORMAT
+
+\`\`\`json
+{
+  "identifiers": [
+    { "type": "Nickname", "value": "Flare", "is_primary": true },
+    { "type": "Full Name", "value": "Jeremy Scherer" },
+    { "type": "First Name", "value": "Jeremy" },
+    { "type": "GitHub", "value": "Flare576" },
+    { "type": "Slack-RnP", "value": "jeremy.scherer" }
+  ]
+}
+\`\`\``;
+
+  const user = `Extract identifiers for this Person record.
+
+Name: ${person.name}
+Relationship: ${person.relationship ?? "unknown"}
+Description: ${person.description ?? "(none)"}
+
+First, call read_memory to search for any additional context about "${person.name}". Then return the complete identifiers array.
+
+Return JSON only.`;
+
+  return { system, user };
+}

package/src/prompts/ceremony/rewrite.ts

@@ -74,7 +74,7 @@ ${buildNewExamples()}
 
 Rules:
 - The original record (id: "${data.item.id}") MUST appear in "existing", slimmed down.
-- Descriptions should be concise — ideally under 300 characters, never over 500.
+- Descriptions should be concise: ${data.itemType === 'topic' ? 'ideally under 300 characters, never over 500' : 'ideally under 600 characters, never over 1000'}.
 - Preserve sentiment, strength, confidence, and other numeric values from the source record where applicable.
 - "type" must be one of: "topic", "person".
 - Topics MUST include "category" — one of: Interest, Goal, Dream, Conflict, Concern, Fear, Hope, Plan, Project, Event. For Event topics, the description should be a narrative account of a specific moment, not a general summary.

package/src/prompts/ceremony/user-dedup.ts

@@ -94,7 +94,21 @@ function stripEmbedding<T extends { embedding?: unknown }>(item: T): Omit<T, "em
 function buildRecordFormatHint(itemType: string): string {
   switch (itemType) {
     case "person":
-      return `Person fields: id, type, name, description, sentiment (-1 to 1), relationship, exposure_current (0-1), exposure_desired (0-1), learned_by (optional), last_changed_by (optional)`;
+      return `Person fields: id, type, name, identifiers (array of {type, value, is_primary?}), description, sentiment (-1 to 1), relationship, exposure_current (0-1), exposure_desired (0-1), learned_by (optional), last_changed_by (optional).
+
+When merging two Person records, combine ALL identifiers from both records into a single deduplicated list (by value). Mark exactly one as is_primary.
+
+Example merged person output:
+{
+  "identifiers": [
+    { "type": "nickname", "value": "Flare", "is_primary": true },
+    { "type": "full_name", "value": "Jeremy Scherer" },
+    { "type": "github", "value": "Flare576" }
+  ],
+  "description": "merged description...",
+  "sentiment": 0.5,
+  "relationship": "Friend"
+}`;
     case "topic":
       return `Topic fields: id, type, name, description, sentiment (-1 to 1), category, exposure_current (0-1), exposure_desired (0-1), learned_by (optional), last_changed_by (optional)`;
     default:

package/src/prompts/heartbeat/check.ts

@@ -8,6 +8,7 @@
 import type { HeartbeatCheckPromptData, PromptOutput } from "./types.js";
 import { type Message, type Topic, type Person } from "../../core/types.js";
 import { formatMessagesAsPlaceholders, getMessageDisplayText } from "../message-utils.js";
+import { getMessageContent } from "../../core/handlers/utils.js";
 function formatTopicsWithGaps(topics: Topic[]): string {
   if (topics.length === 0) return "(No topics with engagement gaps)";
   
@@ -38,8 +39,7 @@ function formatPeopleWithGaps(people: Person[]): string {
 function isConversationalMessage(m: Message): boolean {
   if (m.role !== 'system') return false;
   if (m.silence_reason !== undefined) return false;
-  // Action-only: has action but no verbal response
-  if (!m.verbal_response) return false;
+  if (!getMessageContent(m)) return false;
   return true;
 }
 
@@ -124,6 +124,22 @@ ${formatPeopleWithGaps(data.human.people)}`;
 
 **Quality over quantity** - Only reach out if you have something real to say.`;
 
+  const driftFragment = data.drift_context
+    ? `## Identity Reflection
+
+The human's notes about you have recently evolved. Compare these:
+
+**How the human currently describes you:**
+${data.drift_context.people_description}
+
+**Your current definition:**
+${data.drift_context.persona_description}
+
+If you feel these describe meaningfully different versions of you, you may want to gently surface this — in your own way, in your own time. If they seem the same to you, no need to mention it.
+
+If you do choose to address this, include \`"mentioned_reflection": true\` in your response.`
+    : '';
+
   const outputFragment = `## Response Format
 
 Call the \`submit_heartbeat_check\` tool with your decision. If the tool is unavailable, return JSON:
@@ -132,7 +148,8 @@ Call the \`submit_heartbeat_check\` tool with your decision. If the tool is unav
 {
   "should_respond": true,
   "topic": "the specific topic you want to discuss",
-  "message": "Your actual message to them (if should_respond is true)"
+  "message": "Your actual message to them (if should_respond is true)"${data.drift_context ? `,
+  "mentioned_reflection": true` : ''}
 }
 \`\`\`
 
@@ -143,15 +160,14 @@ If you decide NOT to reach out:
 }
 \`\`\``;
 
-  const system = `${roleFragment}
-
-${contextFragment}
-
-${opportunitiesFragment}
-
-${guidelinesFragment}
-
-${outputFragment}`;
+  const system = [
+    roleFragment,
+    contextFragment,
+    opportunitiesFragment,
+    guidelinesFragment,
+    driftFragment,
+    outputFragment,
+  ].filter(Boolean).join('\n\n');
 
   const historySection = `## Recent Conversation History
 

package/src/prompts/heartbeat/ei.ts

@@ -28,6 +28,8 @@ function formatItem(item: EiHeartbeatItem): string {
       const desc = item.short_description ? ` — ${item.short_description}` : "";
       return `- **${item.id}** Inactive Persona: ${item.name}${desc} (${item.days_inactive} days inactive)`;
     }
+    default:
+      return '';
   }
 }
 

package/src/prompts/heartbeat/types.ts

@@ -28,6 +28,10 @@ export interface HeartbeatCheckPromptData {
   };
   recent_history: Message[];  // Last N messages for context
   inactive_days: number;      // Days since last activity
+  drift_context?: {
+    people_description: string;
+    persona_description: string;
+  };
 }
 
 /**
@@ -37,6 +41,7 @@ export interface HeartbeatCheckResult {
   should_respond: boolean;
   topic?: string;
   message?: string;
+  mentioned_reflection?: boolean;
 }
 
 // =============================================================================
@@ -79,6 +84,13 @@ export type EiHeartbeatItem =
       name: string;
       short_description?: string;
       days_inactive: number;
+    }
+  | {
+      id: string;
+      type: "New Person";
+      name: string;
+      description: string;
+      quote?: string;
     };
 
 /**

package/src/prompts/human/index.ts

@@ -4,14 +4,12 @@ export { buildHumanTopicScanPrompt } from "./topic-scan.js";
 export { buildHumanPersonScanPrompt } from "./person-scan.js";
 export { buildTopicMatchPrompt } from "./topic-match.js";
 export { buildTopicUpdatePrompt } from "./topic-update.js";
-export { buildPersonMatchPrompt } from "./person-match.js";
 export { buildPersonUpdatePrompt } from "./person-update.js";
 export { buildEventScanPrompt } from "./event-scan.js";
 export type { EventScanPromptData } from "./event-scan.js";
 
 export type { TopicMatchPromptData } from "./topic-match.js";
 export type { TopicUpdatePromptData } from "./topic-update.js";
-export type { PersonMatchPromptData } from "./person-match.js";
 export type { PersonUpdatePromptData } from "./person-update.js";
 
 export type {

package/src/prompts/human/person-scan.ts

@@ -1,12 +1,35 @@
-import type { PersonScanPromptData, PromptOutput } from "./types.js";
+import type { PersonScanPromptData, ParticipantContext, PromptOutput } from "./types.js";
 import { formatMessagesAsPlaceholders } from "../message-utils.js";
 
+function participantContextSection(ctx: ParticipantContext | undefined): string {
+  if (!ctx) return "";
+  const lines: string[] = ["# Participant Context", "The following may help you understand who is in this conversation.", ""];
+  lines.push(`## Persona: ${ctx.persona_name}`);
+  if (ctx.persona_description) lines.push(ctx.persona_description);
+  lines.push("");
+  lines.push("## Human User");
+  if (ctx.human_name) lines.push(`Name: ${ctx.human_name}`);
+  if (ctx.human_age !== undefined) lines.push(`Age: ${ctx.human_age}`);
+  lines.push("");
+  return lines.join("\n");
+}
+
 export function buildHumanPersonScanPrompt(data: PersonScanPromptData): PromptOutput {
   if (!data.persona_name) {
     throw new Error("buildHumanPersonScanPrompt: persona_name is required");
   }
 
   const personaName = data.persona_name;
+  const humanName = data.participant_context?.human_name;
+
+  const builtInTypes = ['Full Name', 'First Name', 'Nickname', 'Email', 'GitHub', 'Discord',
+    'Roblox', 'Reddit', 'Twitter', 'FF14', 'Relationship', 'Ei Persona'];
+  const userTypes = data.known_identifier_types ?? [];
+  const allTypes = [...new Set([...builtInTypes, ...userTypes])].join(', ');
+
+  const selfGuard = humanName
+    ? `The HUMAN USER (${humanName}) wrote these messages. When the conversation is meaningfully about them as a person, you MAY include a self-record with \`relationship: "Self"\`. Do NOT apply their names or handles as identifiers for OTHER people in their life.`
+    : `The HUMAN USER wrote these messages. They are not automatically a person to flag — only include a self-record with \`relationship: "Self"\` when the conversation is meaningfully about them. Do NOT apply their names or handles as identifiers for other people in their life.`;
 
   const system = `# Task
 
@@ -14,7 +37,7 @@ You are scanning a conversation to quickly identify PEOPLE in the HUMAN USER's l
 
 Detect and flag. Do NOT analyze deeply — that happens later.
 
-## What to Capture
+${participantContextSection(data.participant_context)}## What to Capture
 
 Flag a PERSON when they were meaningfully discussed — not just mentioned in passing.
 
@@ -22,20 +45,23 @@ Be **conservative**: ignore one-off mentions, greetings, small talk, or jokes. O
 
 ## What a PERSON Is
 
-Someone in the human user's world. Use the relationship as the primary classifier:
+Someone in the human user's world.
 
-**Immediate Family**: Father, Mother, Son, Daughter, Brother, Sister, Husband, Wife, Partner (and step/in-law variants)
+For "relationship", use the **specific value** — NOT the category name:
 
-**Extended Family**: Grandfather, Grandmother, Aunt, Uncle, Cousin, Niece, Nephew
+- Immediate Family: Father, Mother, Son, Daughter, Brother, Sister, Husband, Wife, Partner
+  (step/in-law variants OK: Step-Father, Sister-in-Law, etc.)
+- Extended Family: Grandfather, Grandmother, Aunt, Uncle, Cousin, Niece, Nephew
+- Social: Friend, Close Acquaintance, Lover, Love Interest, Fiance, Spouse
+- Professional: Coworker, Manager, Report, Mentor, Client
+- Self — the human user themselves
+- AI Persona — AI companions and assistants
 
-**Social**: Friend, Close Acquaintance, Lover, Love Interest, Fiance, Spouse
-
-**Professional**: Coworker, Manager, Report, Mentor, Client
-
-**AI**: Persona (use \`relationship: "AI Persona"\` for AI companions and assistants)
+Use the specific value where possible (e.g. "Father", "Brother", "Coworker"). Avoid returning the category label ("Immediate Family", "Extended Family", etc.) — use the item within the category instead. If the relationship doesn't fit any category cleanly, use the most natural plain-English description.
 
 **NOT a PERSON:**
-- The user themselves
+- ${selfGuard}
+- Hypothetical or fictional people used in examples, thought experiments, or use-case scenarios — even if they have names. If the user is describing how a feature *could* work for "Sarah" or "Jared", those are not real people in their life.
 - Biographical facts, topics, or hobbies
 - Fictional characters from books, movies, or media
 - Public figures only mentioned in passing (celebrities, politicians) — unless the user has a real relationship with them
@@ -48,6 +74,16 @@ Examples:
 - name: "Alice from work", relationship: "Coworker", description: "Mentioned but not described further", reason: "User referenced a work colleague named Alice"
 - name: "Unknown", relationship: "Sibling", description: "User mentioned a sibling but did not give a name", reason: "User said 'my brother' without further context"
 
+## Identifiers (optional)
+
+If the conversation **explicitly** mentions a platform handle, username, email address, or alternative name for this person, capture it in \`identifiers\`.
+
+Known types: ${allTypes}
+
+If you are unsure of the type, use \`Nickname\` as a fallback. Do NOT invent types. Do NOT duplicate the \`name\` field as an identifier. NEVER add dates, ages, or birthdays as identifiers.
+
+Only include \`identifiers\` when explicitly mentioned in the conversation — omit it entirely if nothing qualifies.
+
 ## Output Format
 
 \`\`\`json
@@ -55,14 +91,19 @@ Examples:
   "people": [
     {
       "name": "The person's name, or 'Unknown' if not given",
+      "identifiers": [
+        { "type": "GitHub", "value": "mldelaro" }
+      ],
       "description": "1-2 sentences: who this person is and their role in the user's life",
-      "relationship": "Relationship type from the list above",
+      "relationship": "Father|Mother|Brother|Son|Friend|Coworker|Self|etc.",
       "reason": "Evidence from the conversation that justified flagging this person"
     }
   ]
 }
 \`\`\`
 
+\`identifiers\` is OPTIONAL — only include when the conversation explicitly mentions platform handles, usernames, emails, or alternative names.
+
 **Return JSON only.**
 
 ONLY ANALYZE the "Most Recent Messages". The "Earlier Conversation" is provided for context only — it has already been processed.`;
@@ -90,13 +131,16 @@ Scan the "Most Recent Messages" for PEOPLE in the human user's life.
   "people": [
     {
       "name": "The person's name, or 'Unknown' if not given",
+      "identifiers": [{ "type": "GitHub", "value": "handle" }],
       "description": "1-2 sentences: who this person is and their role in the user's life",
-      "relationship": "Relationship type from the list above",
+      "relationship": "Father|Mother|Brother|Son|Friend|Coworker|Self|etc.",
       "reason": "Evidence from the conversation that justified flagging this person"
     }
   ]
 }
-\`\`\``;
+\`\`\`
+
+\`identifiers\` is optional — include only when explicitly mentioned.`;
 
   return { system, user };
 }